Influence of temperature on the biaxial strength of cemented carbides with different microstructures

ACCEPTED MANUSCRIPT 1 Influence of temperature on the biaxial strength of cemented carbides with different microstructures E. Chicardi1*, R.Bermejo2, F. J. Gotor3, L.Llanes4 and Y. Torres1. 1 Departamento de Ingeniería y Ciencia de los Materiales y del Transporte, Escuela Politécnica de Sevilla (EPS). Universidad de Sevilla (US), Virgen de Africa 7, 41011 Sevilla, Spain. 2 Institut fuer Strukturund Funktionskeramik, Montanuniversitaet Leoben, 8700 Leoben, Austria. 3 Instituto de Ciencia de Materiales de Sevilla (US-CSIC), 41092 Sevilla, Spain. 4 Departament de Ciència dels Materials i Enginyeria Metallúrgica, Universitat Politècnica de Catalunya, 08028 Barcelona, Spain. * E-mail: [email protected] Abstract. The effect of the temperature on the mechanical strength of WC-Co cemented carbides with different microstructures (grain size and binder content) was evaluated. Biaxial flexural tests were performed on three cemented carbide grades at 600°C using the ball-on-three-balls (B3B) method. Results were interpreted by Weibull statistics and compared to biaxial strength results at room temperature. A detailed fractographic analysis, supported by Linear Elastic Fracture Mechanics, was performed to differentiate the nature and size of critical defects and the mechanism responsible for the fracture. A significant decrease in the mechanical strength (around 30%) was observed at 600ºC for all grades of cemented carbides. This fact was ascribed to the change in the critical flaw population from sub-surface (at room temperature) to surface defects, associated with the selective oxidation of Co. Additionally, an estimation of the fracture toughness at 600°C was attempted for the three cemented carbides, based upon the B3B strength results, the corresponding number of the tested specimens fragments and the macroscopic area of the B3B fracture surfaces. The fracture toughness was not ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 2 affected by the temperature, at least up to 600ºC. In addition, the good agreement with the Single Edge Notch Beam toughness data suggests the possibility of employing this approach for fracture toughness evaluation of brittle materials under different testing conditions. Keywords: Cemented carbides; temperature; Weibull statistics, biaxial flexural strength; Ball on three Balls. 1. Introduction WC-Co cemented carbides, also known as hard metals, are ceramic-metal composite materials with exceptional combination of hardness, strength, toughness, elastic modulus and wear resistance [1, 2]. Such remarkable mechanical properties result from the synergic effect of their constituent phases, i.e., a hard brittle tungsten carbide (WC) and a soft ductile metallic phase (Co) [3]. The microstructure of WC-Co cemented carbides are usually faceted WC grains, with rectangular or triangular morphology, embedded in a Co-based alloy acting as binder [4]. Other minor secondary carbides (TiC, TaC, NbC, Mo2C, etc.) and metallic elements (Ni, V, Cr, etc.) are added to modulate the microstructure, the chemical stability and the mechanical behavior of these materials [5-7]. Moreover, the use of fine WC particles is also another option to improve their mechanical properties [8]. Cemented carbides are widely used as cutting tools, among other industrial applications. As the cutting speed increases, the operating conditions become more demanding in terms of thermo-mechanical stresses that often endanger their reliability in service. In this regard, cemented carbides exhibit low oxidation resistance and damage tolerance at the high temperatures that can be reached on the tool/chip interface ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 3 while in service [1-4]. It has been reported that the formation of defects at high temperature as a result of oxidation and the growth of the pre-existing flaws are the main causes of the loss of stiffness, mechanical strength and fracture toughness during service [9-12]. These detrimental effects begin to be significant at temperatures above 600ºC [13], and in the 800-900ºC range, depending on the binder content, hard metals reach full oxidation [14]. Acchar et al. [8] found a large loss in mechanical strength from 600ºC and Fantozzi et al. [15] reported that the Young`s modulus and the mechanical strength decreased with temperature, which was particularly important beyond 500ºC. However, a small increase in fracture toughness has been reported between 600-700ºC associated with elastic-plastic transition [15, 16], although this effect is only temporary since from 700ºC-800ºC the fracture toughness diminishes again because of oxidation [15]. Moreover, the mechanical behavior of cemented carbides at high temperature not only depends on the working atmosphere (oxygen partial pressure) and the exposure time, but also on the nature, content and particle size of the constituent phases [9, 14, 17] as well as the inherent flaws associated with the manufacturing process, such as pores, ceramic agglomerates, coarse ceramic particles, eta phases, etc. [18, 19]. At room temperature, cemented carbides show defect controlling fracture strength, which is associated with the size of the largest (or critical) defect in the material. The type of intrinsic defects in these materials ranges from pores to non-metallic inclusions, as well as abnormally large carbides or binderless carbide agglomerates, among other microstructural heterogeneities [20]. This can differ from specimen to specimen (or component to component) [18, 19]. Therefore, strength cannot be described uniquely by a single number, but as a distribution function (related to the defect size distribution), where a probability of failure can be defined as the probability that fracture occurs at a ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 4 stress equal to or lower than a given value, according to Weibull statistics [21, 22]. The probability of failure increases with the magnitude of the load (i.e.  ) and the size of the specimens (i.e. V). The size effect on strength is the most prominent consequence of the statistical behavior of strength in brittle materials, i.e., large specimens have higher probability of failure than smaller specimens under the same load applied. In a previous work, the strength distribution of three cemented carbide grades with different microstructure was evaluated and compared under uniaxial (four-point bend, 4PB) and biaxial (ball-on-three-balls, B3B) bending at room temperature [23]. Characteristic strengths ranging from 2200 MPa to 3000 MPa were measured under 4PB, while higher values between 3300 MPa and 4200 MPa were obtained under B3B. This could be explained according to the size effect on strength in brittle materials, as predicted by the Weibull theory; extrapolation of the 4PB results for effective surface showed good agreement with the B3B experimental values. A higher Weibull modulus was measured in the WC-Co grade with pronounced R-curve under both 4PB and B3B, associated with the subcritical growth of initial defects. The question arises, whether a similar behavior is to be expected at elevated temperatures for the three grades. The aim of this work was to evaluate the influence of temperature on the strength distribution of WC-Co cemented carbides with different microstructures, i.e., grain size and binder mean free path. Biaxial flexural tests were performed at 600°C using the B3B method and the results were compared to those at room temperature. A detailed fractographic analysis was performed to interpret the mechanism responsible for the fracture. Additionally, the fracture toughness of the tested cemented carbides was ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 5 estimated based upon the strength data, the corresponding number of fragments of the tested specimens, and the macroscopic area of the fracture surfaces. 2. Experimental 2.1. Materials Three WC-Co cemented carbide, named as 16F, 16M and 27C, according to a previous published work [3] and with different composition and/or microstructural parameters, were studied. In table 1, the mass percentage of cobalt, the WC average particle size, the binder mean free path and the Vickers hardness of the three different cemented carbides are shown, as characterized in previous work [24]. 2.2. Evaluation of the mechanical strength The mechanical strength of 16F, 16M and 27C cemented carbides was evaluated by biaxial flexural tests using the B3B method [25, 26]. This method is widely used for brittle materials (e.g. glasses, ceramics) [27-29], and has also been applied to cemented carbides [23, 30, 31], and even metal alloys [32], due to some advantages compared to the most common standardized tests, such as 3-point and 4-point bending. Specifically, the main advantages include ease preparation of test samples, high tolerance for some out of flatness of the disc and/or misalignment, much lower friction than in the other common tests, possibility of using small test specimens (even with irregular shape and closer to the real workpiece geometry), absence of tensile loaded edges, among others [23, 25, 27, 28, 33]. The main disadvantage of the B3B method is that the load is concentrated in a very small location of the sample and, thus, larger critical defects may not be sampled during the tests. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 6 Four alumina balls (3 mm of diameter) were used for the tests; three balls placed on the top and the fourth ball centered under the disc specimen. This geometric arrangement during loading guarantees a well-defined three‒point contact. With this configuration, a biaxial tensile stress state at the midpoint of the disc surface opposite the loading ball (down surface) is generated during loading, which is used for the biaxial strength evaluation. More details about the test and a schematic representation can be found elsewhere [23, 33]. At least ten discs (13 mm of diameter and 0.7 – 1.3 mm of thickness) of each cemented carbide grade were tested. The surfaces were carefully ground and polished (final step with suspension diamond powder of 3 µm) to prevent surface damage and ensure the same surface quality in all specimens. The B3B tests were carried out by an electromechanical universal testing machine (Model 5505, Instron Ltd.) in ambient air, under load control at a rate of 100 N/s, coupled to a split furnace (Energon S.L) with refractory Molybdenum clamps (m.p. = 2623 ºC). The assembly of the device and disc was initially made at room temperature. Then, a 10 N pre-load was applied in order to guarantee the contact between the four balls and the disc during the heating step. The temperature was increased up to 600 ºC (2 h were necessary to reach the temperature), always controlling and holding the imposed 10 N pre-load. In the B3B test, the fracture load, F (in N), is used to calculate the maximum biaxial flexural strength, σmax, according to the following equation [25]: max 2   F ft (1) where t (in mm) is the disc thickness and f a non-dimensional factor, which depends on the specimen geometry and position of the balls, the Poisson's ratio of the material, and ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 7 the special features of the load transfer from the clamps to the disc. In this work, the values of f ranged between 1.1 and 1.5 as determined from the equation by Börger, Supancic and Danzer [25], due to the different specimen thicknesses. The fracture strength was statistically analyzed as a function of stress, which was related to a certain failure probability, Pf, using a two-parameter Weibull distribution. Assuming an homogeneous crack-size frequency density function [34], the following failure probability function, Pf, can be defined [22]:   f 00 , 1 exp             m V PV V (2) where σ0 and m are the characteristic strength and the Weibull modulus, respectively. The Weibull modulus, m, describes the scatter of the strength data, and the characteristic strength,  0, is the stress at which, for specimens of volume V=V0, the failure probability is: Pf(  0, V0) = 1-exp(-1) ≈ 63%. The determination of the Weibull distribution was performed according to the standard EN 843-5 [35]. In addition, the 90% confidence intervals (CI) were calculated for  0 and m, which represent the range where the true Weibull parameters (i.e. from the parent distribution) can be found with a 90% probability and that reflect the influence of the sampling procedure. They were determined according to the standard EN 843-5 [35]. The fracture surfaces of the tested discs were analyzed by Scanning Electron Microscopy (SEM) using a Hitachi S-4800 SEM-Field Emission Gun microscope and the Image Analysis software Image-Pro Plus 6.2 (Media Cybernetics Inc., USA). These fractographic analyses allow the determination and evaluation of the nature and size of ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 8 the flaws causing the failure, and (in many cases) the (micro)-mechanism that may has been responsible for the fracture. In addition, the WC, Co and O contents in the WC-Co cemented carbides were measured using an energy dispersive X-ray spectrometry (XEDS) detector coupled to the SEM microscope. Finally, the measurements of the total area of the fracture surface, Sf, and the number of disc fragments associated with the fracture process, Ni, was carried out on a representative number of discs for each grade of cemented carbides using a Nikon Epiphot optical microscope coupled with a Jenoptik Progres C3 camera and the analysis software Image-Pro Plus 6.2. 3. Results and discussion 3.1. Influence of temperature on the biaxial strength Fig. 1 shows the probability of failure versus the corresponding failure stress (Weibull plots) for the three commercial cemented carbides grades (16F, 16M and 27C) at room temperature (RT) and 600°C. The calculated σ0 and m values are presented in table 2, together with the corresponding 90% CI. A detrimental effect of temperature on the mechanical strength of the three cemented carbides is evidenced in Fig. 1 (see characteristic values in table 2). The decrease of approx. 30% in strength at 600ºC compared to RT may be mainly attributed to oxidation phenomena, as was directly observed at first sight on the surface of the discs after testing. As reported in the literature, the formation of oxidation products begins to be appreciable at 600°C [10]. In fact, previous works have shown a significant reduction of the mechanical strength of cemented carbides only after 10 min at 700ºC [36]. Therefore, it is not surprising that an ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 9 oxidation process was associated with the testing procedure, since the samples were heated during 2 h to reach the testing temperature of 600°C. In addition to the loss of mechanical strength at 600ºC, an apparent increase of the Weibull modulus, m, was observed for the 16F and 16M grades (the mean value of m increasing from 8 to 10, and 7 to 13, for the two grades respectively). A higher m may be associated with the subcritical growth of preexisting processing flaws, thus yielding a higher amount of defects of similar (larger) size. Another possibility might be the appearance of new flaws with a more homogeneous size distribution and location. In the case of grade 27C, it seemed not to be an influence of the temperature on the Weibull modulus (m=10 at RT versus m=9 at 600°C). It is worthy pointing out that the 90% CI for m do overlap for the three WC-Co grades. In order to reinforce the hypothesis mentioned above for 16F and 16M materials on the change of flaw distributions with temperature, more experiments would be required that may narrow the 90% CI, and thus reveal statistically significance differences in m. This aspect will be discussed in more detail in the following section, supported by fractographic considerations. The effect of the microstructure on the mechanical strength of the cemented carbides is illustrated in Fig. 2a and 2b at RT and at 600ºC, respectively. It can be observed how the influence of the microstructure on the probability of failure of the specimens was different for both evaluated temperatures. At RT and lower mechanical strength values, corresponding to larger inherent flaws (Fig. 2a), all three grades seem to have same behavior. In addition, a deviation from the expected Weibull behavior was observed in the 16M and 27C grades. In these two grades, the obtained curves using 2parameter Weibull distribution do not fit the strength data; there seems to be an apparent “threshold stress”, at approx. 2800 MPa, below which the material does not fail. However, for smaller inherent flaws (higher mechanical strength values in Fig. 2a), the ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 16 critical defects due to the selective oxidation of the binder phase. In addition, the fracture toughness of the three cemented carbides was not affected by the temperature, at least up to 600ºC. The viability of the B3B biaxial flexural test for testing WC-Co was demonstrated at elevated temperatures, enabling the observation of oxidation phenomena and their consequence on strength by revealing a new critical (surface) flaw population compared to sub-surface flaws found at room temperature. In addition, a direct relationship between the numbers of broken fragments after B3B testing, the mechanical strength and the total fracture surface area was observed. This interesting protocol developed by Yanaba and Hayasaki [46] and corroborated here for the three cemented carbides studied, opens the possibility to generate diagrams for different brittle materials, connecting their fragment numbers after mechanical testing with their fracture toughness. Of special interest is the application of this procedure when the fracture toughness testing may become difficult. Acknowledgments This work was supported by the Andalusian government (Spain) under the excellence project no. P12-TEP-2622. Funding for this investigation was also partly supplied by the Spanish MINECO/FEDER under Grant No. MAT2015-70780-C4-3-P. The authors want to thank the technician J. Pinto and M. Sánchez for their assistance with mechanical testing and microstructural characterization. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 17 References [1] Tarragó JM, Jiménez-Piqué E, Schneider L, Casellas D, Torres Y, Llanes L. FIB/FESEM experimental and analytical assessment of R-curve behavior of WC–Co cemented carbides. Materials Science and Engineering: A. 2015;645:142-9. [2] Mari D, Bolognini S, Feusier G, Viatte T, Benoit W. Experimental strategy to study the mechanical behaviour of hardmetals for cutting tools. International Journal of Refractory Metals & Hard Materials. 1999;17:209-25. [3] Llanes L, Torres Y, Anglada M. On the fatigue crack growth behavior of WC-Co cemented carbides: kinetics description, microstructural effects and fatigue sensitivity. Acta Materialia. 2002;50:2381-93. [4] Lavergne O, Robaut F, Hodaj F, Allibert CH. Mechanism of solid-state dissolution of WC in Co-based solutions. Acta Materialia. 2002;50:1683-92. [5] Juhani K, Pirso J, Viljus M, Letunovits S, Tarraste M. The Influence of Cr3C2 and VC as Alloying Additives on the Microstructure and Properties of Reactive Sintered WC-Co Cermets. Materials Science-Medziagotyra. 2012;18:79-83. [6] Lee HR, Kim DJ, Hwang NM, Kim DY. Role of vanadium carbide additive during sintering of WC-Co: Mechanism of grain growth inhibition. Journal of the American Ceramic Society. 2003;86:152-4. [7] Zhang L, Xie M-w, Cheng X, Nan Q, Wang Z, Feng Y-p. Micro characteristics of binder phases in WC-Co cemented carbides with Cr-V and Cr-V-RE additives. International Journal of Refractory Metals & Hard Materials. 2013;36:211-9. [8] Acchar W, Gomes UU, Kaysser WA, Goring J. Strength Degradation of a Tungsten Carbide-Cobalt Composite at Elevated Temperatures. Materials Characterization. 1999;43:27-32. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 18 [9] Basu SN, Sarin VK. Oxidation behavior of WC-Co. Materials Science and Engineering a-Structural Materials Properties Microstructure and Processing. 1996;209:206-12. [10] Voitovich VB, Sverdel VV, Voitovich RF, Golovko EI. Oxidation of WC-Co, WCNi and WC-Co-Ni hard metals in the temperature range 500-800 degrees C. Int J Refract Met Hard Mater. 1996;14:289-95. [11] Arenas F, Ribas N, Brito J, Albornoz A, Cho SA. Examination of oxidation products in WC-Co hard metals. Bristol: Iop Publishing Ltd; 1998. [12] Lee G. Oxidation Behavior of WC-Co Hardmetal. Joural of Korean Power Metallurgy Instit. 2004;11:111-7. [13] Huang S, Xiong J, Guo Z, Wan W, Tang L, Zhong H, et al. Oxidation of WC-TiCTaC-Co hard materials at relatively low temperature. International Journal of Refractory Metals & Hard Materials. 2015;48:134-40. [14] Babutina TE, Uvarova IV. Effect of the conditions of oxidation of lump wastes of WC-Co(Ni) hard alloys on the dispersity of the oxidized mixtures. Powder Metallurgy and Metal Ceramics. 1997;36:107-10. [15] Fantozzi G, Mohand H, Orange G. High temperature mechanical behaviour of WC6 wt.% Co cemented carbide. Science of hard materials. 1986;20:699-712. [16] Milman YV, Luyckx S, Goncharuck VA, Northrop JT. Results from bending tests on submicron and micron WC-Co grades at elevated temperatures. International Journal of Refractory Metals & Hard Materials. 2002;20:71-9. [17] Bhaumik SK, Balasubramaniam R, Upadhyaya GS, Vaidya ML. oxidation behavior of hard and binder phase modified WC-10Co cemented carbides. Journal of Materials Science Letters. 1992;11:1457-9. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 19 [18] Davidge RW. Mechanical Behaviour of Ceramics. Cambridge: Cambridge University Press; 1980. [19] Danzer R. Ceramics: mechanical performance and lifetime prediction. Oxford: Pergamon Press; 1994. [20] Fischmeister H. Development and Present Status of the Science and Technology of Hard Materials. In: R.K. Viswanadham DJRaJG, editor. Proc of ICSHM1 - 1st Int Conf on the Science of Hard Materials. New York: Plenum Press; 1981. p. 1-45. [21] Danzer R. A general strength distribution function for brittle materials. J Eur Ceram Soc. 1992;10:461-72. [22] Weibull W. A statistical distribution function of wide applicability. J Appl Mech. 1951;18:253. [23] Torres Y, Bermejo R, Gotor FJ, Chicardi E, Llanes L. Analysis on the mechanical strength of WC-Co cemented carbides under uniaxial and biaxial bending. Materials & Design. 2014;55:851-6. [24] Torres Y, Casellas D, Anglada M, Llanes L. Fracture toughness evaluation of hardmetals: influence of testing procedure. Int J Refract Met Hard Mater. 2001;19:2734. [25] Börger A, Supancic P, Danzer R. The ball on three balls test for strength testing of brittle discs: Stress distribution in the disc. Journal of the European Ceramic Society. 2002;22:1425-36. [26] Danzer R, Harrer W, Supancic P, Lube T, Wang Z, Börger A. The ball on three balls test-Strength and failure analysis of different materials. Journal of the European Ceramic Society. 2007;27:1481-5. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 20 [27] Bermejo R, Supancic P, Aldrian F, Danzer R. Experimental approach to assess the effect of metallization on the strength of functional ceramic components. Scripta Materialia. 2012;66:546-9. [28] Bermejo R, Supancic P, Krautgasser C, Morrell R, Danzer R. Subcritical crack growth in Low Temperature Co-fired Ceramics under biaxial loading. Engineering Fracture Mechanics. 2013;100:108-21. [29] Wagner R, Harrer W, Danzer R. Application of the Ball on Three Balls Test in the Development of a High Strength Partially Stabilised Zirconia Ceramic. Cfi-Ceramic Forum International. 2009;86:E50-E3. [30] Chicardi E, Torres Y, Cordoba JM, Hvizdos P, Gotor FJ. Effect of tantalum content on the microstructure and mechanical behavior of cermets based on (TixTa1x)(C0.5N0.5) solid solutions. Materials & Design. 2014;53:435-44. [31] Chicardi E, Torres Y, Córdoba JM, Sayagués MJ, Rodríguez JA, Gotor FJ. Effect of sintering time on the microstructure and mechanical properties of (Ti,Ta)(C,N)-based cermets. International Journal of Refractory Metals and Hard Materials. 2013. [32] Benavente-Martinez E, Devesa F, Amigo V. Mechanical characterization of Ti-Nb alloys by ball on three balls test. Revista De Metalurgia. 2010;46:19-25. [33] Börger A, Supancic P, Danzer R. The ball on three balls test for strength testing of brittle discs: Part II: Analysis of possible errors in the strength determination. Journal of the European Ceramic Society. 2004;24:2917-28. [34] Bermejo R, Danzer R. 2.09 - Mechanical Characterization of Ceramics: Designing with Brittle Materials A2 - Sarin, Vinod K. Comprehensive Hard Materials. Oxford: Elsevier; 2014. p. 285-98. [35] EN 843-5. Advanced Technical Ceramics - Monolithic Ceramics - Mechanical Tests at Room Temperature - Part 5: Statistical Analysis. EN 843-51997. p. 40. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 21 [36] Casas B, Ramis X, Anglada M, Salla JM, Llanes L. Oxidation-induced strength degradation of WC-Co hardmetals. International Journal of Refractory Metals & Hard Materials. 2001;19:303-9. [37] Irwin GR. Fracture. In: Flügge S, editor. Handbuch der Physik. Berlin: SpringerVerlag; 1958. p. 551-89. [38] Johannesson B, Warren R. Subcritical crack growth and plastic deformation in the fracture of hard metals. Materials Science and Engineering: A. 1988;105-106:353-61. [39] Sheikh S, M'Saoubi R, Flasar P, Schwind M, Persson T, Yang J, et al. Fracture toughness of cemented carbides: Testing method and microstructural effects. International Journal of Refractory Metals & Hard Materials. 2015;49:153-60. [40] Wang G, Wang YT, Liu YH, Pan MX, Zhao DQ, Wang WH. Evolution of nanoscale morphology on fracture surface of brittle metallic glass. Applied Physics Letters. 2006;89:3. [41] Riedle J, Gumbsch P, Fischmeister HF. Cleavage anisotropy in tungsten single crystals. Physical Review Letters. 1996;76:3594-7. [42] Garcia-Rosales C, Lopez-Ruiz P, Alvarez-Martin S, Calvo A, Ordas N, Koch F, et al. Oxidation behaviour of bulk W-Cr-Ti alloys prepared by mechanical alloying and HIPing. Fusion Engineering and Design. 2014;89:1611-6. [43] Gu W-H, Jeong YS, Kim K, Kim J-C, Son S-H, Kim S. Thermal oxidation behavior of WC–Co hard metal machining tool tip scraps. Journal of Materials Processing Technology. 2012;212:1250-6. [44] Aristizabal M, Sanchez JM, Rodriguez N, Ibarreta F, Martinez R. Comparison of the oxidation behaviour of WC–Co and WC–Ni–Co–Cr cemented carbides. Corrosion Science. 2011;53:2754-60. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 22 [45] Chen L, Yi D, Wang B, Liu H, Wu C. Mechanism of the early stages of oxidation of WC–Co cemented carbides. Corrosion Science. 2016;103:75-87. [46] Yanaba Y, Hayashi K. Relation between fracture surface area of a flexural strength specimen and fracture toughness for WC-10mass%Co cemented carbide and Si3N4 ceramics. Materials Science and Engineering a-Structural Materials Properties Microstructure and Processing. 1996;209:169-74. [47] Danzer R. On the relationship between ceramic strength and the requirements for mechanical design. J Eur Ceram Soc. 2014;34. ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 23 Table 1. Microstructural parameters and mechanical properties of the WC-Co cemented carbides studied. WC-Co grade wt.% Co dwc (µm) Co (µm) HV30 (GPa) 16F 10 0.50 0.25 15.5 ± 2.5 16M 10 1.06 0.30 14.0 ± 2.6 27C 16 1.66 0.76 11.2 ± 0.8 ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 24 Table 2. Influence of the microstructure and the temperature on the Weibull parameters (σ0 and m) for the three cemented carbides studied at room temperature (RT) [23] and at 600 °C. The values shown in square brackets correspond to the 90% confidence intervals for the Weibull parameters. WC-Co grade Characteristic strength, σ0 (MPa) [90% CI] Weibull modulus, m [90% CI] RT 600 °C RT 600 °C 16F 4208 [3888-4567] 2978 [2787-3189] 8 [4-11] 10 [5-13] 16M 3281 [3012-3585] 2258 [2148-2377] 7 [4-10] 13 [7-17] 27C 3341 [3133-3571] 2331 [2156-2529] 10 [6-13] 9 [5-12] ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 25 Table 3. Critical flaw sizes, 2ac: (i) estimated at RT and (ii) measured on fractured specimens at 600 °C. Fracture toughness values, KIc, estimated at 600 °C using two different geometric factors: Y=2/ as for sub-surface flaws, and Y=(2/)·21/2 as for surface flaws contacting the surface. WC-Co grade Critical flaw sizes, 2ac (µm) Measured KIc by SENB at RT [24] (MPam1/2) Estimated KIc by LEFM at 600ºC (MPam1/2)** Estimated KIc by LEFM at 600ºC (MPam1/2)*** @RT (estimated)* @600ºC (measured) 16F 6–9 8-10 9.2 ± 0.6 7.1 ± 0.4 9.9 ± 0.4 16M 14–19 19-22 10.5 ± 0.8 8.1 ± 0.4 11.4 ± 0.4 27C 27–35 30-32 14.7 ± 1.0 10.1 ± 1.1 14.1 ± 1.1 * 2ac values at RT estimated using LEFM from measured σ0 [23], KIc [24] and Y=(2/) (as for sub-surface flaws [47]) ** KIc values @600°C estimated using LEFM from measured 2ac, σ0 and Y=(2/) (as for sub-surface flaws [47]) *** KIc values @600°C estimated using LEFM from measured 2ac, σ0 and Y=(2/)·21/2 (as for surface flaws contacting the surface [47]) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 32 3 µm 3 µm 27C_RT 27C_600ºC W W Co Co O O Figure 5 ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 33 W Co O 27C_600ºC Figure 6 30 µm ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 34 Figure 7 ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 35 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 0 5 10 15 20 Number of pieces 0 (GPa) 16F_RT 16F_600ºC 16M_RT 16M_600ºC 27C_RT 27C_600ºC Figure 8 ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 36 0 1 2 3 4 5 6 7 8 9 10 0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 3.2 16F Linear Fit 16M Linear Fit 27C Linear Fit y = 356·x y = 329·x y = 378·x 0 (GPa) 16F 16M 27C Sf 1/2 (mm) Figure 9 ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 37 HIGHLIGHTS  The effect of the temperature on the WC-Co mechanical strength was determined.  Biaxial flexural tests were performed at 600°C by the ball-on-three-balls method.  A significant decrease in the mechanical strength (30%) was observed at 600ºC.  This fact was ascribed to the change of sub-surface to surface critical flaws.  Additionally, the fracture toughness was not affected by the temperature. ACCEPTED MANUSCRIPT