V Congreso Nacional de Pulvimetalurgia Girona 2015. 1, 2 y 3 de julio XXX-XXX

MICROSTRUCTURAL INFLUENCE ON THE FATIGUE BEHAVIOUR OF CEMENTED CARBIDES

J. M. Tarragó1,2, L. Schneider3, I. Al-Dawery3, L. Llanes1,2

1 CIEFMA, Departament de Ciència dels Materials i Enginyeria Metal∙lúrgica, Universitat Politècnica de Catalunya, 08028, Barcelona

2 CRnE, Centre de Recerca en Nanoenginyeria, Universitat Politècnica de Catalunya, 08028, Barcelona

3 SANDVIK HYPERION, Coventry CV4 0XG, UKe-mail del autor: jose.maria.tarrago@upc.edu

RESUMEN

En esta investigación se ha evaluado la influencia de la microestructura en el comportamiento a fatiga de los carburos cementados. Con este propósito, se han seleccionado tres variables microestructurales (tamaño de grano y la naturaleza química y contenido de fase ligante) y se ha estudiado su efecto en el comportamiento de crecimiento de grieta por fatiga (FCG). Los resultados indican una dependencia lineal entre el umbral de fatiga (Kth) y el tamaño de grano debido a la acción de la deflexión de grietas como un mecanismo de aumento de tenacidad inmune a la fatiga. Por otra parte, no se ha observado una clara influencia de la naturaleza química del ligante en el comportamiento a fatiga. Asimismo, se puede establecer el umbral de fatiga como la tenacidad efectiva de los carburos cementados bajo cargas cíclicas debido a la excelente correlación encontrada entre el límite a fatiga y Kth. Se ha realizado un análisis fractográfico detallado de los micromecanismos de daño dentro del ligante y se ha observado que, independientemente de la naturaleza química del ligante, éste tiene un aspecto marcadamente cristalográfico donde se aprecian claramente escalones. La presencia de estos escalones puede ser racionalizada en base a una región de deformación plástica en la punta de la grieta de tamaño similar a la unidad microestructural.

ABSTRACT

In this study, the microstructural influence on the fatigue behaviour of cemented carbides is assessed. With this purpose, three microstructural variables (carbides grain size and binder chemical nature and content) are selected and its effect on the fatigue crack growth (FCG) behaviour investigated. Results indicate a linear dependence between the FCG threshold (Kth) and the grain size due to the action of crack deflection as a fatigue immune toughening mechanisms. On the other hand, binder chemical nature is not found to have a significant influence on FCG rates. In addition, FCG threshold is stated as the effective toughness due to the excellent correlation found between determined fatigue limit strength and Kth. A detailed fractographic study of damage micromechanisms has been conducted and, regardless of the binder chemical nature, well-defined faceted crystallographic features have been observed within the binder. Such features are rationalized on the basis of a cyclic plastic zone in front of the crack tip of similar size to the microstructural unity.

Palabras clave (Keywords): cemented carbides, fatigue behaviour, fractography, fatigue crack growth threshold

1

Microstructural influence on the fatigue behaviour of cemented carbides

1. INTRODUCTION

A better understanding of service degradation phenomena in hardmetals is required for industrial manufacturers if material performance and lifetime of tools and components are to be improved. Among them, fatigue is a relevant one since cemented carbides are commonly used in applications involving high cyclic stresses, and is associated with premature and unexpected failure [1]. The more relevant scientific and technical advances on fatigue of cemented carbides are concentrated in the last two decades. Among them it is worth underlining the findings reported by Sockel's [2–5] and Llanes's [6–8] groups, developed following different testing approaches: applied stress–fatigue life (S–N curves) and fatigue crack growth (FCG) tests, respectively. Schleinkofer and co-workers documented a strong strength degradation of cemented carbides under cyclic loads mainly related to fatigue degradation localized in the ductile binder phase [2–4]. Furthermore, through their systematic work, they were able to identify the subcritical crack growth of preexisting flaws as the controlling stage for fatigue failure in these materials [2–5]. On the other hand, Torres and co-workers proposed the FCG threshold as the effective toughness under cyclic loading [6]. Moreover, they pointed out that fatigue sensitivity of hardmetals is significantly dependent on microstructure, according to the compromising role played by the binder as the main toughening and fatigue susceptible agent in cemented carbides [7]. This fatigue degradation ascribed to the metallic phase, also observed from the experimental S–N data published by Sailer and co-workers [9], has been rationalized – at least partly – on the basis of the cyclic strain-induced fcc to hcp phase transformation within the Co binder [2,3]. On the other hand, one of the major trends in hardmetal industry is focused on finding new binder phases to replace cobalt [10]. Among them nickel has attracted considerable attention as substitutes for cobalt because of its similarity in structure and properties [11], besides its good corrosion resistance [11,12]. However, information on the fatigue behaviour of Ni-base cemented carbides is rather scarce (e.g. Refs. [5,13,14]). In this regard, nickel binder accumulates deformation in the form of slip plus twinning damage mechanisms [5,13,14], but without evidence of phase transformation, as it is the case of Co binder. Thus, it is not clear whether above relationships proposed by Llanes an co-workers [6,7], regarding either fatigue mechanics perspective or microstructural influence on the basis of binder mean free path, may be directly extrapolated to other than Co-base hardmetals. Within this context, one of the main goals of this investigation is to evaluate the influence of the partial or total replacement of Co binder by Nickel on the fatigue behavior of cemented carbides by extrapolating the work performed by Llanes and co-workers on WC-Co hardmetals [6,7] to Ni-containing ones. On the other hand, in a recent work Torres et al. [8] pointed out the effective action of crack deflection as an additional toughening mechanism (in addition to the bridging mechanism at the crack wake) immune to fatigue loads. Therefore, it is also the purpose of this investigation to study the influence of microstructural coarsening on the fatigue crack growth behavior of hardmetals.

2. MATERIALS END EXPERIMENTAL ASPECTS

Six hardmetal grades corresponding to different combinations of binder chemical nature, binder content and carbide grain size were studied. All materials were supplied by Sandvik Hyperion. The designations and key microstructural parameters: binder content (wt.% binder), mean grain size (dWC), carbide contiguity (CWC), and binder mean free path (λbinder) are listed in Table 1. Mean grain size was measured following the linear intercept method by means of FESEM, using a JEOL-7001F unit. Carbide contiguity and binder mean free path were deduced from best-fit equations, attained after compilation and analysis of data published in the literature (e.g. Refs. [15,16]), on the basis of empirical relationships given by Roebuck and Almond [17] but extending them to include the influence of carbide size [18].

2

Microstructural influence on the fatigue behaviour of cemented carbides

Table 1. Microstructural parameters for the studied hardmetal grades.

Grade wt.% binder dWC

(µm) CWCλbinder

(µm)10CoUF 10%Co 0.39 ± 0.19 0.46 ± 0.06 0.16 ± 0.0611CoM 11%Co 1.12 ± 0.71 0.38 ± 0.07 0.42 ± 0.2810CoC 10%Co 2.33 ± 1.38 0.31 ± 0.11 0.68 ± 0.4815CoM 15%Co 1.15 ± 0.92 0.30 ± 0.07 0.55 ± 0.469NiF 9%Ni 0.77 ± 0.49 0.45 ± 0.08 0.27 ± 0.18

10CoNiM 8%Co-2%Ni 1.04 ± 0.83 0.41 ± 0.08 0.36 ± 0.29

Mechanical characterization includes hardness (HV30), flexural strength (σ r), fracture toughness (KIc) and fatigue-crack-growth (FCG) parameters. Hardness was measured using a Vickers diamond pyramidal indenter under a load of 294N and the reported value is the average of five measurements. In all the other cases, testing was conducted using a four-point bending fully articulated test jig with inner and outer spans of 20 mm and 40 mm, respectively. Flexural strength tests were performed on an Instron 8511 servohydraulic machine at room temperature. At least 15 specimens (45 mm × 4 mm × 3 mm) were tested per grade. The surface (which was later subjected to the maximum tensile loads) was polished to a mirror-like finish and the edges were chamfered to reduce their effect as stress raisers. Results were analyzed using Weibull statistics. Fracture toughness and FCG parameters were determined using 45 mm × 10 mm × 5 mm single-edge pre-cracked notch beam (SEPNB) specimens with a notch length-to-specimen width ratio of 0.3. Compressive cyclic loads were induced in the notched beams to nucleate a sharp crack; details are given elsewhere [19]. The sides of the SEPNB specimens were polished to follow stable crack growth using a high-resolution confocal microscope. Fracture toughness was determined by testing the SEPNB specimens to failure at stress intensity-factor load rates of ~2 MPa√m/s. FCG behaviour was assessed for two different load-ratio (R) values, 0.1 and 0.5, using a Rumul Testronic resonant testing machine at load frequencies ~150 Hz. Experimental fatigue limit (‘‘infinite fatigue life’’ defined at 106 cycles) was assessed following the stair-case method for the 11CoM and 10CoC hardmetal grades. Tests were performed using the same resonant testing machine, at load frequencies around 150 Hz and under a load ratio (R) of 0.1.

3. RESULTS AND DISCUSSION

Hardness, flexural strength, Weibull modulus, and fracture toughness of the studied materials are listed in Table 2. The dispersion in flexural strength is relatively small and, accordingly, the corresponding Weibull analysis yields relatively high values for all investigated hardmetals, indicative of high reliability from a structural viewpoint.

Table 2. Hardness, strength parameters and fracture toughness for investigated hardmetals.

Grade HV30(GPa)

σr

(MPa)Weibullmodulus

KIc

(MPam)10CoUF 15.7 ± 0.6 3422 ± 512 11 10.4 ± 0.311CoM 12.8 ± 0.2 3101 ± 102 36 13.9 ± 0.310CoC 11.4 ± 0.2 2489 ± 85 35 15.8 ± 0.315CoM 11.2 ± 0.1 2912 ± 88 39 15.2 ± 0.49NiF 13.4 ± 0.1 3085 ± 182 20 11.5 ± 0.2

10CoNiM 12.3 ± 0.1 2720 ± 198 23 14.2 ± 0.4

FCG rate versus the stress intensity factor (Kmax) plots, including data for the two studied load ratios (0.1 and 0.5), are shown in Figure 1 for the studied cemented carbides. FCG thresholds (Kth), defined at crack growth rates of 10−6 mm/cycle, were attained following a decremental

3

Microstructural influence on the fatigue behaviour of cemented carbides

loading sequence. As previously reported, studied hardmetals exhibit FCG at Kmax values much lower than corresponding fracture toughness levels, and large-power dependences between FCG rates (da/dN) and Kmax.

Figure 1. Determined crack growth rates (da/dN) plotted against Kmax for the investigated cemented carbides determined at two different load ratios of 0.1 and 0.5

Table 3 shows the determined fatigue threshold and fatigue sensitivity [1−(Kth/KIc)] levels for the studied hardmetal grades. FCG parameters are quite similar for grades with alike microstructures (i.e. carbide size and binder content), but different binder chemical nature. On the other hand, coarse-grained materials exhibit higher fatigue threshold values than those determined for medium-grained ones. This improvement is related to a more effective crack-deflection [8] mechanism, as microstructure gets coarser, which remains operative under cyclic loading (i.e. toughening discerned under monotonic loading is not inhibited or degraded by fatigue). In Figure 2 the FCG threshold is plotted as a function of the carbide mean grain size for the investigated hardmetals and for the two studied load ratios. It is found that the fatigue threshold (Kth) correlates linearly with the carbide mean grain size (dWC).

Table 3. Hardness, strength parameters and fracture toughness for investigated hardmetals.

Grade Kth (R=0.1)(MPa*m1/2)

Kth (R=0.5)(MPa*m1/2)

1- Kth/KIc

(R=0.1)1- Kth/KIc

(R=0.1)10CoUF 6.2 7.2 0.40 0.3111CoM 7.6 8.6 0.45 0.3810CoC 8.3 10.4 0.47 0.3415CoM 7.6 9.1 0.50 0.409NiF 7.2 8.2 0.37 0.29

10CoNiM 7.8 9 0.45 0.37

4

Microstructural influence on the fatigue behaviour of cemented carbides

Figure 2. Fatigue threshold as a function of the mean grain size for the cemented grades studied here.

Llanes et al. [7] investigated the dependence of fatigue sensitivity of WC–Co hardmetals as a function of the binder mean free path and applied load ratio. A similar study has been attempted and in Figure 3 the fatigue sensitivity exhibited by the studied cemented carbides is plotted against the binder mean free path for load ratios of 0.1 and 0.5. It is interesting to note that the fatigue sensitivity linearly increases with the binder mean free path, indicating that as the binder ligaments become larger the material is more susceptible to fatigue. Main reason for that is that the effective toughening given by the metallic phase becomes degraded under the application of cyclic loads [7]. Meanwhile, fatigue sensitivity of Co, Ni and CoNi based cemented carbides follow similar trends. On the other hand, experimental results indicate smaller fatigue sensitivity values for coarse grained grades than that expected from the direct extrapolation of the fatigue sensitivity data obtained for fine and medium grained materials. Once again, the higher relevance of crack deflection mechanism as microstructure gets coarser, should here be recalled for rationalizing such finding [8].

Figure 3. Fatigue sensitivity as a function of binder mean free path for the hardmetals studied here.

Based on the assumption that subcritical crack growth is the controlling stage for fatigue failure in cemented carbides [2–5], Torres and co-workers proposed and successfully validated a FCG threshold–fatigue limit correlation for WC–Co hardmetals within the Linear Elastic Fracture Mechanics (LEFM) framework [6,20]. In doing so, they defined the FCG threshold as the effective toughness under the application of cyclic loads for an infinite life approach. Thus, the fatigue limit (σf) can be deduced from the stress intensity factor threshold of a small non-propagating crack emanating from a defect of critical size, 2acr, through a relationship of type:

5

Microstructural influence on the fatigue behaviour of cemented carbides

σ f =Y ∙K th

√acr

(1)

where Y is a crack geometry factor. Accordingly, fatigue limit (σf) values can be estimated from the relation given by expression (1) under the assumption that failure-controlling flaws under the application of monotonic and cyclic loadings have similar size, geometry and distribution.

σ f

σ r=

K th

K Ic (2)

Aiming to validate the above relationship, fatigue limit values were experimentally determined, following an up-and-down load (staircase) methodology, for two of the hardmetal grades under consideration: 11CoM and 10CoC (Figure 4). An excellent agreement is obtained between predicted – using relationship (2) – and experimentally determined fatigue limits (Table 4), validating then the referred FCG threshold–fatigue limit correlation. A detailed FE-SEM fractographic examination of samples broken under cyclic loads pointed out critical defects of similar nature and geometry to those observed in failed specimens tested under monotonic loading. However, under the application of cyclic loads these defects act as starting locations for subcritical crack growth until they reach a critical size where unstable fracture takes place (Figure 5). As previously reported, stable crack growth region within the binder is characterized by crystallographic step-like features [7,8,21]. These step-like features have been observed in all hardmetals studied here, indicative that they are independent on the binder chemical nature and pointing out the step-like crack morphology to be rather a microstructure size scale effect. In this regard, it is well-known that the transition from the near-threshold regime to the intermediate stage: (i) is accompanied by a noticeable change from a microstructure-sensitive to a microstructure-insensitive fracture behavior; and (ii) occurs when the size of the cyclic plastic zone (rc) becomes comparable to the characteristic microstructural dimension of the material under consideration [22]. When plane stress conditions are satisfied, the size of the cyclicplastic region can be approximated as

rc ≈ 1π ( ∆ K I

2 σ y) (3)

where ΔKI is the applied stress intensity factor range and σy is the yield strength of the material. In cemented carbides, the high effective yield stress exhibited by the constrained binder (between 2 and 4 GPa [16]), together with the relatively low ΔKI values at which stable crack growth takes place (between 4 and 10 MPa·m1/2, Figure 1), yields a submicrometric plastic region ahead the crack tip, whose size is then comparable to the binder mean free path. Under these conditions, microscopic failure modes characterized by localized shear and zig-zag crack paths may be expected, as it is evidenced in this investigation too. Therefore, such features are speculated to be related to a cyclic plastic zone in front of the crack tip of similar size to the microstructural unity [23].

Table 4. Estimated and experimentally determined fatigue limit, in terms of maximum applied stress, for 11CoM and 10CoC hardmetal grades studied.

Grade Estimated fatigue limit(MPa)

Experimentally determined fatigue limit(MPa)

11CoM 1696 ± 70 1603 ± 97

6

Microstructural influence on the fatigue behaviour of cemented carbides

10CoC 1308 ± 53 1250 ± 84

(a) (b)

Figure 4. Example of a microstructure heterogeneity acting as an initial point for fatigue crack growth. Fatigue crystallographic features are observed in the vicinity of the starting critical flaw.

Figure 5. Example of a microstructure heterogeneity acting as an initial point for fatigue crack growth. Fatigue crystallographic features are observed in the vicinity of the starting critical flaw.

4. CONCLUSIONS

The influence of the microstructure on the fatigue behaviour of cemented carbides has been investigated. Based on obtained results the following conclusions may be drawn:

1. FCG threshold of cemented carbides linearly rises when increasing carbide mean grain size, due to a more effective action of the crack deflection mechanism. Therefore, coarse-grained hardmetals are found to be less fatigue sensitive than finer-graded with similar binder mean free path.

7

Microstructural influence on the fatigue behaviour of cemented carbides

2. FCG threshold and fatigue sensitivity are quite similar for grades with alike microstructures (i.e. carbide size and binder content), but different binder chemical nature. This negligible influence of binder chemical nature points out then that ductile Co, Ni and CoNi binders exhibit similar degradation susceptibility under fatigue of toughening mechanisms operative under monotonic loading.

3. A fatigue mechanics analysis allows estimating the fatigue limit of hardmetals on the basis that Kth is the effective toughness under cyclic loads.

4. Stable FCG for the hardmetal studied is characterized by faceted, crystallographic features within the binder, different from the ductile dimples evidenced in the region of unstable propagation. Such fatigue failure mode is postulated to be a direct consequence of the comparable size length scales of microstructure and cyclic plastic zone in front of the crack tip.

5. ACKWNOLEDGMENTS

This work was financially supported by the Spanish Ministerio de Economía y Competitividad (GrantMAT2012-34602). In addition, J.M. Tarragó acknowledges the Ph.D. scholarship received fromthe collaborative Industry-University program between Sandvik Hyperion and Universitat Politècnica de Catalunya.

6. REFERENCES

[1] L. Llanes, M. Anglada, Y. Torres, Fatigue of cemented carbides, in: V.K. Sarin, D. Mari, L. Llanes (Eds.), Comprehensive Hard Materials, Elsevier, 2014: pp. 345–362.

[2] U. Schleinkofer, H.G. Sockel, K. Gorting, W. Heinrich, Microstructural processes during subcritical crack growth in hard metals and cermets under cyclic loads, Mater. Sci. Eng. A. 209 (1996) 103–110.

[3] U. Schleinkofer, H.G. Sockel, K. Gorting, W. Heinrich, Fatigue of hard metals and cermets, Mater. Sci. Eng. A. 209 (1996) 313–317.

[4] U. Schleinkofer, H.G. Sockel, K. Görting, W. Heinrich, Fatigue of Hard Metals and Cermets - New Results and a Better Understanding, Int. J. Refract. Met. Hard Mater. 15 (1997) 103–112.

[5] S. Kursawe, P. Pott, H.G. Sockel, W. Heinrich, M. Wolf, On the influence of binder content and binder composition on the mechanical properties of hardmetals, Int. J. Refract. Met. Hard Mater. 19 (2001) 335–340.

[6] Y. Torres, M. Anglada, L. Llanes, Fatigue mechanics of WC-Co cemented carbides, Int. J. Refract. Met. Hard Mater. 19 (2001) 341–348.

[7] L. Llanes, Y. Torres, M. Anglada, On the fatigue crack growth behavior of WC–Co cemented carbides: kinetics description, microstructural effects and fatigue sensitivity, Acta Mater. 50 (2002) 2381–2393.

8

Microstructural influence on the fatigue behaviour of cemented carbides

[8] Y. Torres, J.M. Tarrago, D. Coureaux, E. Tarrés, B. Roebuck, P. Chan, M. James, B. Liang, M. Tillman, R.K. Viswanadham, K.P. Mingard, A. Mestra, L. Llanes, Fracture and fatigue of rock bit cemented carbides: Mechanics and mechanisms of crack growth resistance under monotonic and cyclic loading, Int. J. Refract. Met. Hard Mater. 45 (2014) 179–188.

[9] T. Sailer, M. Herr, H.-G. Sockel, R. Schulte, H. Feld, L.J. Prakash, Microstructure and mechanical properties of ultrafine-grained hardmetals, Int. J. Refract. Met. Hard Mater. 19 (2001) 553–559.

[10] S. Norgren, J. García, A. Blomqvist, L. Yin, Trends in the P/M hard metal industry, Int. J. Refract. Met. Hard Mater. 48 (2015) 31–45.

[11] V.A. Tracey, Nickel in Hardmetals, Int. J. Refract. Met. Hard Mater. 11 (1992) 137–149.

[12] W.J. Tomlinson, C.R. Linzell, Anodic polarisation and corrosion of cemented carbides with cobalt and nickel binders, J. Mater. Sci. 23 (1988) 914–918.

[13] E.F. Drake, A.D. Krawitz, Fatigue damage in a WC-Nickel cemented carbide composite, Metall. Mater. Trans. A. 12A (1981) 505–513.

[14] C.H. Vasel, A.D. Krawitz, E.F. Drake, E.A. Kenik, Binder deformation in WC-(Co, Ni) cemented carbide composites, Metall. Mater. Trans. A. 16A (1985) 2309–2317.

[15] R.K. Viswanadham, T.S. Sun, E.F. Drake, J. Peck, Quantitative fractography of WC-Co cermets by Auger spectroscopy, J. Mater. Sci. 16 (1981) 1029–1038.

[16] L.S. Sigl, H.F. Fischmeister, On the fracture toughness of cemented carbides, Acta Metall. Mater. 36 (1988) 887–897.

[17] B. Roebuck, E.A. Almond, Deformation and fracture processes and the physical metallurgy of WC-Co hardmetals, Int. Mater. Rev. 33 (1988) 90–110.

[18] D. Coureaux, Comportamiento mecánico de carburos cementados WC-Co: Influencia de la microestructura en la resistencia a la fractura, la sensibilidad a la fatiga y la tolerancia al daño inducido bajo solicitaciones de contacto., Universitat Politècnica de Catalunya, 2012.

[19] Y. Torres, D. Casellas, M. Anglada, L. Llanes, Fracture toughness evaluation of hardmetals: influence of testing procedure, Int. J. Refract. Met. Hard Mater. 19 (2001) 27–34.

[20] Y. Torres, M. Anglada, L. Llanes, Fatigue limit - fatigue crack growth threshold correlation for hardmetals: influence of microstructure, in: A.F. Blom (Ed.), Proceedings 8th International Fatigue Congress, EMAS, Stockholm, 2002.

[21] G. Erling, S. Kursawe, S. Luyck, H.G. Sockel, Stable and unstable fracture surface features in WC-Co, J. Mater. Sci. Lett. 19 (2000) 437–438.

[22] S. Suresh, Fatigue of materials, 2nd ed., Cambridge university press, New York, 1998.

[23] J.M. Tarragó, C. Ferrari, B. Reig, D. Coureaux, L. Schneider, L. Llanes, Mechanics and mechanisms of fatigue in a WC–Ni hardmetal and a comparative study with respect to WC–Co hardmetals, Int. J. Fatigue. 70 (2015) 252–257.

9