Room temperature fatigue crack growth in cemented carbides

Room temperature fatigue crack growth in cemented carbides

Materials Science and Engineering, 83 (1986) L7-L10 L7 Letter Room temperature fatigue crack growth in cemented carbides S. SURESH and L. A. SYLVA ...

456KB Sizes 0 Downloads 12 Views

Materials Science and Engineering, 83 (1986) L7-L10

L7

Letter

Room temperature fatigue crack growth in cemented carbides S. SURESH and L. A. SYLVA Division of Engineering, Brown University, Providence, RI 02912 (U.S.A.)

(Received April 24, 1986) ABSTRACT

Fully compressive cyclic loads applied to notched plates o f a WC-Co composite are found to result in stable Mode I fatigue crack growth at room temperature. Possible mechanisms leading to this compression fatigue effect are contemplated. A novel technique for fracture testing o f cermets is recommended whereby fracture toughness values can be determined in bending or in tension after fatigue pre-cracking notched specimens in uniaxial cyclic compression.

Ceramic-metal composites such as cemented carbides represent a class of materials where the high strength and hardness of ceramics and the ductility, impact resistance and toughness of the metallic binder material are optimized for applications in a variety of engineering situations. The fracture properties of these materials have been the subject of much research work in the past. However, there have been only a few prior studies of the influence of cyclic loads on the fracture behavior in ceramic-metal systems [ 1-4] and the fatigue of cermets has remained a relatively unexplored area of research. The resistance of these materials to crack growth under cyclic compressive stresses is a topic of particular interest in many engineering applications and, to date, this has n o t been studied. The objective of this Letter is to d o c u m e n t and discuss the results and significance of the experimental observation that fatigue crack propagation occurs in notched plates of ceramic-metal composites subjected to fully compressive 0025-5416/86/$3.50

far-field cyclic loads at r o o m temperature. The material examined in this study is a commercially available WC-25wt.%Co composite (grade K-90 from Kennametal Inc., Latrobe, PA). This material also contains trace amounts of TaC, TiC, nickel and iron. The r o o m temperature mechanical properties are as follows: compressive strength, 3482 MPa; modulus of elasticity, 444 GPa; mean transverse rupture strength, 2861 MPa. Fatigue crack growth experiments were carried o u t on single-edge-notched specimens machined to the following dimensions: width W = 12.69 mm; height H = 12.69 mm; thickness B = 3.25 mm; notch-length-to-specimen-width ratio 5/w = 0.33; notch r o o t radius p ~ 0.5 mm. Fully compressive cyclic stresses were applied at a constant stress amplitude IAa ~ ] of 865 MPa, a frequency v of 4 Hz and a load ratio R of 10. The load ratio is defined as the ratio of the minimum load to the maximum load of the fatigue cycle. The specimen was located between t w o parallel plates aligned with guide pins and fatigued in an electroservohydraulic machine in the laboratory environment (temperature, 23 °C; relative humidity, 40%). The crack length, as a function of the number of compression cycles, was monitored with the aid of an optical microscope. Figure 1 is an optical micrograph which shows the morphology of a m o d e I fatigue crack propagated from the notch r o o t under fully compressive cyclic loads. This micrograph was obtained after the crack had completely arrested. (Some secondary cracking was also noticed at the sharp corners at the notch root; the length of the secondary (nonpropagating) flaw was insignificant in comparison with that of the main crack shown in Fig. 1). The variation in crack length, as a function of the number of compression cycles, is plotted in Fig. 2. The broken line represents the average values of crack length a from several measurements corresponding to the fixed cyclic loads. The m a x i m u m uncertainty in the measured crack length is indicated by © Elsevier Sequoia/Printed in The Netherlands

L8

OJ

.

Fig. 1. Optical micrograph showing the mode I fatigue crack propagated from the notch root during the application of fully compressive cyclic loads. The arrows indicate the direction of compression loading. Fig. 3. Micrograph showing the path of the fatigue crack with reference to the microstructure. 1.0

~oe 06

c% 0 4

//

o F- 0 2 O0

0

Oil

'

012

'

0!5

'

041

'

05

NUMBER OF COMPRESSION CYCLES ( x 106)

Fig. 2. The variation in crack length a (measured from the notch tip) as a function of the compression cycles. The scatter band reflects the differences in crack length on the two side surfaces of the specimen.

the scatter band on the discrete data points. Observations on the fracture surface after completion of the test revealed a uniform crack front through the center thickness of the specimen; some non-uniformity in the crack front was observed only at the slowest growth rates (at high crack length levels) just prior to crack arrest. As the fatigue crack tip advances away from the notch root, closure effects become dominant in the wake of the crack tip. An important consequence of crack closure is that the fraction of the fatigue cycle during which the crack tip remains open under far-field compression progressively decreases with increasing crack length [ 5]. Therefore, the fatigue crack arrests completely after stable growth over a certain distance (about 0.6 m m in Fig. 2). As the region of damage at

the crack tip is always surrounded by material elastically strained in compression, the crack growth process is non-catastrophic. The path of the fatigue crack through the microstructure of the WC-Co composite is illustrated in the micrograph in Fig. 3 which was obtained by examining the polished-and-etched surface of the specimen. This figure shows that crack propagation occurs primarily along the interface between the brittle carbide and the ductile binder material. Some evidence of growth within the metallic phase can also be observed. It should be noted that there is no cracking whatsoever within the brittle phase. Possible mechanisms underlying this phenomenon are examined in the following. It has long been known that cyclic compressive loads applied to notched components of metallic materials can result in the propagation of a mode I fatigue crack from the root of the notch [5-8]. This phenomenon is promoted by the existence of a region of residual tensile stress in the vicinity of the notch tip as a result of cyclic plasticity during unloading from the maximum compression condition. The fatigue cracks which easily initiate from the notch tip under residual tension propagate at a progressively decreasing rate before arresting completely. Although the occurrence of this phenomenon in metals has long been recognized, only recently have

L9 extensive theoretical and experimental studies been undertaken to elucidate the effects of stress state, loading variables, crack closure and microstructure on the mechanics and mechanisms of crack growth under far-field cyclic compression [ 5, 7, 8]. As the residual tensile stress induced by reversed plasticity (i.e. the cyclic plastic zone) at the r o o t of the notch is the primary "driving f o r c e " for crack advance under fully compressive applied loads in metals, the significance of this p h e n o m e n o n had n o t been explored in brittle solids where the extent of plasticity is very limited at low temperatures. However, Ewart and Suresh [9, 10] recently reported experimental results which clearly demonstrated that r o o m temperature fatigue crack propagation can occur even in brittle materials subject to fully compressive cyclic loads. They observed stable non-catastrophic (macroscopically) m o d e I crack growth in notched specimens of single-phase polycrystaUine aluminum oxide loaded in cyclic compression. The initially rapid rate of crack growth progressively decelerated, resulting eventually in crack arrest [9]. The mechanisms underlying this effect, however, are different from those observed in metals. On a microscopic scale, fatigue frature is p r o m o t e d by the nucleation of grain b o u n d a r y microcracks in the vicinity of the notch r o o t during the application of far-field compressive stresses. Residual tensile stresses are induced at the notch tip if the microcracks within the notch-tip damage zone remain open during unloading from the maxim u m far-field compression load. As shown by Brockenbrough and Suresh, it is this region of residual tensile stresses at the notch tip which promotes Mode I fatigue under far-field cyclic compression. The sliding along a grain boundary microcrack promotes a tensile crack at an adjacent grain boundary. (This intrinsic micromechanism of failure is well d o c u m e n t e d for brittle solids such as rocks and concrete subject to quasi-static compression [9].) The coalescence of such microcracks into a macrocrack is aided by the shear stresses (on the grain boundaries) arising from the far-field cyclic compressive loads and by the residual tensile and shear stresses arising from differential ~hermal contraction between the randomly oriented grains in non-cubic ceramics [9, 10]. The occurrence of a predominantly intergranular fracture m o d e and the absence of this

effect in oriented crystals of sapphire appear to lend support to such a hypothesis. Furthermore, the growth of stable fatigue cracks in polycrystalline alumina cycled under far-field compression in vacuo clearly indicates that this p h e n o m e n o n is induced by an intrinsic mechanical failure process and not by any stress corrosion effects [10]. In ceramic-metal composites, such as cemented carbides, fatigue crack growth under imposed compression can occur by an interaction between the mechanisms observed in metals and ceramics. On the basis of the above discussion, some important factors which influence this p h e n o m e n o n in cermets can readily be identified. (a) Far-field fatigue compressive loads create residual tensile stresses as a consequence of cyclic plasticity in the metallic binder. (In a metallic material exhibiting an elasticperfectly plastic deformation behavior, residual tensile stresses of magnitude equal to the flow stress are present at the notch tip over a distance of the order of one-quarter of the monotonic compressive plastic zone [6-8]). (b) A damage zone at the notch tip, characterized b y a population of microcracks at the carbide-binder interface, can produce a local reduction in stiffness. If a fraction of the microcracks remains open during the unloading portion of the compression cycle, residual tensile stresses will be generated at the notch tip [11]. (c) Residual stresses are produced at the carbiae-binder interface because of thermal c o n t r a c t i o n mismatch induced by the large difference between the coefficients of thermal expansion for the two materials (e.g. the coefficients of thermal expansion for WC and cobalt differ by as much as a factor of 10}. (c) The shear stresses and sliding along the carbide-binder interface (exerted by the farfield cyclic compressive loads) can p r o m o t e tensile cracks at an adjacent interface [9, 10]. Although we have presented in this Letter the first experimental evidence o f fatigue crack growth in compression only in a cermet with 25 wt.% metallic phase, subsequent work in other ceramic-metal composites (e.g. WC6wt.%Co and TiC-12wt.%Ni) has also successfully demonstrated the feasibility of obtaining stable r o o m temperature crack growth under imposed compression cycles [12]. While this Letter is intended to d o c u m e n t the possibility

L10

of stable crack growth under far-field cyclic compression, further research is needed to elucidate the mechanisms of this effect as a function of applied load range, details of the microstructure including the relative amounts of the carbide and the binder, geometry of the notch, load ratio and environment. The measurement of a "valid" fracture toughness value in cemented carbides has traditionally been somewhat restricted by the experimental difficulty of "pre-cracking" the fracture specimen (in tensile fatigue). In this work, we have established a procedure for introducing controlled mode I (non-catastrophic) fatigue flaws in ceramic-metal composites even at low temperatures. The capability for producing stable non-catastrophic flaws (sharp fatigue pre-cracks) using this compression technique in brittle materials at room temperature provides unique possibilities for improved measurements of fracture properties than can be achieved using currently available methods. The introduction of a self-arresting fatigue crack under cyclic compression also enables a quick and easy measurement (in a single specimen) of the threshold stress intensity factor at which stress corrosion, creep or (tensile) fatigue crack growth begins. Here, subsequent to pre-cracking in compression, the specimen can be loaded in static tension or bending (for stress corrosion or creep tests) or in cyclic tension or bending (for fatigue tests) at stress intensity levels lower than the expected threshold values. The loads are incremented until crack growth is detected and the entire crack growth curve, from threshold to final fracture, can be obtained in a single specimen. As the compression fatigue crack propagates by progressively decreasing near-tip damage and arrests naturally, the maximum amount of near-tip damage at the point of arrest may not influence subsequent crack growth in tension. We note that further detailed work on the mechanics and micromechanisms of this phenomenon are required before its attractive features as a pre-cracking

technique can be successfully exploited in a variety of fracture studies in ceramic-metal composites.

ACKNOWLEDGMENTS

This work was supported by the National Science Foundation under Grant NSF-ENG8451092. The authors thank H. Stanton for his help during the course of this work.

REFERENCES 1 E. A. Almond and B. Roebuck, Met. Technol., 7 (1980) 83. 2 R. Fry, Fatigue crack growth behavior of WC-Co hardmetal alloys, Ph.D. Thesis, University of Witwatersrand, 1982. 3 A. G. Evans and M. Linzer, Int. J. Fract., 12 (1976) 217. 4 I. Johansson, G. Persson and R. Hiltscher, Powder Metall., 13 (1970) 449. 5 S. Suresh, Eng. Fract. Mech., 21 (3) (1985) 453463. 6 R. P. Hubbard, J. Basic Eng., 91 (1969) 625. 7 D. K. Holm, A. F. Blom and S. Suresh, Eng. Fract. Mech., 23 (6) (1986), (1986) 1097. 8 S. Suresh, T. Christman and C, Bull, Crack initiation and growth under far-field cyclic compression: theory, experiments and applications, in R. O. Ritchie and J. Lankford (eds.), Small Fatigue Cracks, Metallurgical Society of AIME, Warrendale, PA, 1986. 9 L. Ewart and S. Suresh, J. Mater. Sci. Lett., 5 (1986), 774. 10 L. Ewart and S. Suresh, Crack propagation in ceramics under cyclic loads, J. Mater. Sci., 22 (1987), in the press. 11 J. R. Brockenbrough and S. Suresh, Constitutive behaviour of a microcracking brittle solid in cyclic compression, Brown University Report NSF-ENG-FRACT-1, August 1986. 12 R. Godse, unpublished research in progress, Brown University, Providence, RI, 1986,