The McGraw-Hill Companies 34__Elements of Fracture Mechanics Taste 2.4 The table of Problem 7 Crack length a (nim) Compliance C (miN) Critical load P(N) 32 33.5 x 10-6 94.3 2 68.9 x 10° 78.8 50 121.0 x 10-* 64.0 58 191.0 x 10 535 67 294.0 x 10-6 44.0 REFERENCES 2.1 Kanninen, M.E. and Popelar, C . H. (1985). Advanced Fracture Mechanics, Oxford University Press, New York. 2.2 Gdoutos, EE. (1993). Fracture Mechanics—An Introduction, Kluwer Academic Publishers, Dordrecht. 2.3 Broek, D. (1982). Elementary Engineering Fracture Mechanics, Martinus Nijhoff Publishers, The Hague. 2.4 Meguid, S.A. (1989). Engineering Fracture Mechanics, Elsevier Applied Science, London. 2.5 Barsom, J. M. and Rolfe, S.T. (1987). Fracture and Fatigue Control in Structures, Prentice- Hall, Inc., Englewood Cliffs, New Jersey. 2.6 Hertzberg, R. W. (1989). Deformation and Fracture Mechanics of Engineering Materials, John Wiley and Sons, New York 2.7 Knott, JF. (1973). Fundamentals of Fracture Mechanics, John Wiley and Sons, New York. 2.8 Griffith, A.A. (1921). The Phenomena of Rupture and Flow in Solids, Philosophical ‘Transactions of the Royal Society of London, A 221, pp 163-169. 2.9 Griffith, A.A. (1924). The Theory of Rupture, Proceedings of the First International Conference of Applied Mechanics, Delft 2.10 Inglis, C. E. (1913). Stress in a Plate due to the Presence of Cracks and Sharp Corners, Transactions of the Institute of Naval Architects, 55, pp. 219-241 2.11 Irwin, G. R. (1948). Fracture Dynamics: Fracturing of Metals, American Society for Metals, Cleveland, pp. 147-166. 2.12 Orowan, E. (1948). Fracture and Strength of Solids, Reports on Progress in Physics, XII, p- 185. 2.13 Anderson, T.L. (2004). Fracture Mechanics: Fundamentals and Applications, CRC, Press-Book. 2.14 Sanford, R. J. (2003). Principles of Fracture Mechanics, Prentice Hall, Upper Saddle River. 2.15 Janssen, M,, Zuidema, J. & Wanhill, RJ.H. (2004). Fracture Mechanics, Spon Press, Abingdon. 2.16 Ramesh, K. (2007). e-Book on Engineering Fracture Mechanics, ITT Madras, URL: http//apm.iitm.ac.in/smlab /kramesh/book_4_htm.The McGraw-Hill companies Stress Intensity Factor nature is intelligent because it is too complicated to be called anything else. Deepak Chopra 3.1 InTRopuCTION 3.1.1 Why Should Investigations be Closer to the Crack Tip? Today, thinking big often implies focusing on minute details. Huge amounts of atomic energy is generated, both for peaceful and destructive purposes, by manipulating atoms which are invisible even under powerful microscopes. Similarly, in order to understand how crystalline materials (eg., metals) deform plastically, we should understand the behavior of dislocations, which are imperfections and look like threads. Dislocations too are very small in diameter, but can be viewed under a powerful microscope. A crack front exists within a material, like a line running from one region of the body to another. The vicinity of a crack tip offers interesting information as the magnitudes of stress components are extremely high. When we want to break a wooden stick by bending it we make a notch. The notch creates high stresses, which in turn make the notch tip move rather easily. Knowing the stress or displacement field in the vicinity of a crack tip is very useful. A material scientist may devise ways to develop new materials which can diffuse high stresses at the crack tip. A designer may modify some features such as notches, cutouts, keyways, etc., to minimize stresses. An experimentalist can think of methods of characterizing cracks by measuring stresses or strains near the crack tip. One of the biggest advantages is that stress analysis leads to define a parameter, stress intensity factor (SIF) to characterize a crack. In comparison to energy release rate, SIF is simpler for a designer and easier for laboratory measurements, so as to determine material properties. 3.1.2 Linear Elastic Fracture Mechanics (LEFM) In a brittle material, the material remains elastic even at the crack tip where stresses are high. Most engineering materials do not fall in this category. Diamonds are known to be quite brittleMee ae 36 Elements of Fracture Mechanics (clastic) even in the vicinity of a crack tip. Other brittle materials like window panes are known to have some anelastic deformation close to the crack tip. This chapter develops analysis of brittle materials only; the analysis does not account for plastic deformation close to the crack tip. A natural question arises—when most materials are not brittle, why should we make detailed analysis of brittle materials? Obviously, analysis of brittle materials is far simpler than the analysis of a material having a plastic zone at the crack tip. The presence of a plastic zone means that two kinds of stress-strain behaviors should be incorporated, plastic behavior inside the plastic zone and elastic behavior outside it. At the same time, itis difficult to evaluate the interface between the two zones. Furthermore, the material behavior inside the plastic zone is complex to model and in some materials, especially those which exhibit pronounced Bauschinger effect, the material behavior is extremely difficult to account for. Therefore, for the analysis of elastic-plastic fracture mechanics (EPFM), problems are usually solved through numerical analysis, By confining our attention to elastic (or brittle) materials, we would be able to obtain closed form solutions to many problems. Also, we will learn how to deal with the singularity (infinite stresses at the crack tip) involved. There is another advantage in developing solutions of elastic crack problems. In many real life cases where the plastic zone size is quite small in comparison to the crack length, the contribution of the plastic zone in an elastic analysis may be neglected. That is, if stress fields in such cases are determined for purely elastic and elastic-plastic cases separately, the difference between the two is small enough to be considered negligible. A large number of engineering problems of practical applications fall in this category and consequently elastic analysis is good enough. This leads to an important subfield—linear elastic fracture mechanics (LEEM), where only elastic analysis is carried out to determine stress and displacement fields near a crack tip with characterizing parameters like the SIF. It is worth mentioning here that the energy release rate (G) has also been formulated for LEFM. 3.2 STRESS AND DisPLACEMENT FIELDs IN IsoTROPIC ELASTIC MATERIALS On what parameters does the stress field depend in the vicinity of the crack tip? It, of course, depends on the external load, generally denoted by the far field stress o.'The name, far field, may mislead some readers; it refers to stress at points of the body where the influence of the high stresses generated at the crack tip is negligible. The second important parameter, as discussed in Chapter 2, is the crack length 2. We also realize that the stress field varies near the tip significantly from point to point and one must specify the coordinates at which stress components are to be determined. Generally, polar coordinates (r, @) are employed for describing the location of the point. Further, the stress field depends upon geometry i.c., whether the crack is at the centre of the body, at the edge, or at the off-centre. Under geometrical considerations, the effect of the overall size of the specimen with respect to the crack length would also be included. Let us express stress component g;; in the vicinity of crack tip as 6,,=/lo,a, r, 8, geometry) where i, j are the suffixes representing various stress components like 64, G22, 0;2, Gay, ete. There is a separate relation (function f) for each stress component. In engineering applications, most of the hardware components are in the form of plates. Commonly used sections like angles, I-beams, channels, rectangular tubes, etc., are made of flat plate sections. Even in a circular tube with a crack having its length much smaller than the diameterThe McGraw-Hill Companies Stress Intensity Factor _37 of the tube, the region around the crack tip may be regarded asa flat plate. Thus, a study of crack in a flat plate encompasses many engineering applications. In most problems of this book, x;-axis is chosen along the crack length, x,-axis along the normal of the plate and x,-axis in the plane of the plate and normal to the crack length [Fig. 3.1(a)]. Also, we have chosen to denote stress components with two suffixes because that is a better way of representing the stress tensor. Consequently, we would be using symbol ¢ for shear stress components also; the suffixes will only differentiate shear components from normal components. Further, we prefer to use suffixes 1, 2 and 3 in place of x, y and z, as used in many books. Fig. 3.1 Axes with respect to the crack in a plate Solving equations of solid mechanics to obtain expressions for 6, in the vicinity of the crack tip is mathematically involved and, therefore, only the results would be presented and discussed in this section. The derivations would be presented in Sections 3.4 and 3.5. To present results, we have chosen the common case of a flat plate with a crack of length 2a and far field stress os shown in Fig. 3.2(a). The stress field at a general point H near the crack tip for isotropic and linear elastic material in the flat plate for this Mode I case is Fig. 3.2 (a) Infinite plate with a crack of length 2a subjected to a far field stress 0, and (b) definition of stress components at point HBee ma Red 38 Elements of Fracture Mechanics ora? oo 0. 30] n= Gani 68 a|1-sin 5 sin | (3-1a) n ; 0 = EO cos 2 [t+sin 8 sin 30 (3.1b) Qnr) 2 22) o(nay* 9 88 6,2 ZAI sin © cos © cos G10) Qn) 2 ZOD For a thin plate, other stress components are negligible. In case of a thick plate 033 = V(o1, + G22) where vis the Poisson's Ratio of the material; the other two stress components (13, 633) are negligible Itis clear from these equations that each stress component is proportional to the far field stress a. The crack length appears under a square root and, therefore, its influence on stress components is also prominent, but not to the extent of o. The distance (r) between the crack tip and the point plays the most important role. It sits in the denominator under a square root sign. If r becomes very small, the stress components, specially Gy, goes up steeply, so much so that for r > 0, dy: tends to be infinite. Such solutions are called singular. In this case, it is known as square root singularity. In some special cases of fracture mechanics, other kinds of singularities are encountered. Displacement field for a plane strain near the crack tip for Mode I of Fig. 3.2(a) is given by mae, 2 y= sea (5) cos$f 1-27 + sin? | (3.2a) wae ny Wwe ty = sea (5) sin [2-20 4 00? | (3.2b) uw Qe us =0 (3.20) where # is the shear modulus. The above equations do not contain any singularity, because displacement is finite near the crack tip. Mathematically, displacement components are obtained from stress equations by converting stress components fo strain components and then integrating the resulting expressions. During integration, the square root singularity disappears and displacement components turn out to be proportional to the square root of the distance r. However, one should note here that the Eqs (3.2a-c) are valid only in the close vicinity of the crack tip. 3.3 Stress INTENSITY FACTOR In the engineering field, a problem with two variables is much more difficult to solve than a problem of one variable. A question may be raised here—can two independent variables be combined to form a new independent variable? If the answer is yes, the solution to the problem is likely to become much simpler. In the case of a wave propagation, solution even along one dimension becomes complex because two variables, space (x) and time (t), are involved. In such wave problems, we usually combine two variables to form a new variable, (x cl), where cis theThe McGraw-Hill companies Stress Intensity Factor 39 velocity of the sound. There are many other similar cases such as linear kinetic energy combining two variables to one, linear momentum combining two variables to one, Reynolds Number in fluid mechanics combining four independent variables to one, and Sommerfeld Number in journal bearing combining five variables into one. If we look carefully into Eqs (3.1) and (3.2), we find that for a given geometry there are two main variables, the far field stress o and the crack length a. Furthermore, in all the equations of stress and displacement, ganda coexist as o Va . Can this product be called by a different variable? Now with several decades of research work, we find that it is advantageous to do so. This credit goes to Irwin [3.1], who defined the new variable, stress intensity factor, and used the symbol K after the name of his collaborator Kies [3.2]. He defined K as K,= 6 (na)? (3.3) There is no reason to have 7 in the above definition. It was included in the expression because of some historical reasons which shall be explained in the next chapter. However, the stress intensity factor K; is formally defined as Ky = (229)! 6) (r, 0= 0) as r—30 (3.4) This definition can be checked easily by substituting Eq. (3.1b) in the formal definition [Eq, (3.4)], the resulting expression will be the same as of Eq. (3.3). We realize that, for stress or displacement fields, magnitude of o or a is immaterial as long as @(za)"’? is same. This means a small crack length in a plate with high far field stress is equivalent to a large crack length with small far field stress, provided K remains same. This combination of @ and a to form a new variable is regarded as a breakthrough in the field of fracture mechanics. For Modes I-III, the stress intensity factor is written as K,, Ky and Kyy respectively, with subscript in Roman numbers. The stress and displacement Eqs (3.1) and (3.2) may now be written in terms of the stress intensity factor. For Mode I problems of plane strain, they become a 6, 38 = —"1 5 cos 2} 1-sin = sin 5i one ghigeseet[t-sint sn] 5 or 0. 38 a= ag cos S| 1¢sin % sin cae afi crt fiein in] oxy Kr_-sin 2 cos ® cos @ (3.50) 1 pee ZO ZOD °) 1/2 y= A(r) cos Ab = 2v +sin? $] (35d) a an 2 2 v2 Ar) sin s[2- 2v + cos? 5] (3.5e) races 2 2 The stress intensity factor elegantly characterizes a crack, similar to energy release rate, G, developed in Chapter 2, Equations (3.5a-e) need to be modified for bodies which are of finite dimensions or where the crack tip is close to one of the free edges of the component. Closed form expression for stress intensity factor is available only for simple cases and, therefore, determiningBe me Red 40 _ Elements of Fracture Mechanics the stress intensity factor becomes a challenge for many practical cases. These days, numerical techniques are widely used for this very purpose. However, for a body with a crack and known boundary conditions, once the stress intensity factor is determined, the crack is characterized for a designer and then, he can predict whether a crack in the work-component is likely to grow or not Stress and displacement equations for the center-cracked body are similar for other modes. For Mode Il in plane strain and far field stress 612 = t (Fig. 3.3) with Ky = ra, we have For Mode Ill and far field stress 0, = t (Fig. 3.4) with Ky = tra, we have Gy, = Gx = G33 = Gy, =0The McGraw-Hill companies Stress Intensity Factor 41 Fig. 3.4 A centre-crack in an infinite plate loaded in Mode III 3.4 BACKGROUND FOR MATHEMATICAL ANALYSIS In the previous section, stress and displacement fields in the vicinity of a crack tip were presented without providing proofs. This section would develop the mathernatical base and the following section would solve problems of all the three modes for simple geometries. Although the field equations of solid mechanics are quite well-known, they would be presented for plane problems for the sake of completeness. The solution to fracture problems would be obtained by solving field equations of solid mechanics. Of course, proper boundary conditions would be incorporated, including zero traction conditions (free surface) of the cracked faces. Toa beginner, some parts of this section may look strange and may be a bit difficult in the first reading because such analysis is not usually seen in other fields of mechanics of solids. A straightforward and simple analysis is not possible because the problem deals with a singularity. We would proceed slowly as if we are trying to reach the centre of a flower bud by taking one petal out gently at a time. This, of course, comes at a risk of annoying some of the readers who are already conversant with this fieldMe ee Re ed 42__ Elements of Fracture Mechanics 3.4.1 Field Equations There are three kinds of field equations [3.3] which are solved for a set of given boundary conditions of a component. They are: (i) equilibrium equations relating to stress components, (ii) strain-displacement relations and (iii) stress-strain relations for the given material of the component. In case the problem is solved with stress or strain components as dependent variables, an additional set of equations must be satisfied to ensure that during the deformation, a continuous body remains continuous. This condition is called compatibility [3.4]. We will develop it because most problems would be solved in stress components. In fact, the compatibility condition provides the governing differential equation for many problems. Equilibrium Equations: Equilibrium equations within a body are developed in any textbook of solid mechanics and, therefore, they will not be derived here. For problems dealing with plates, variation of stress components in the thickness direction is assumed to be negligible. For most fracture mechanics problems, body force is not important and, therefore, we are left with the following two differential equations of equilibrium doy | a6, in , 202 29 3.6a ax 8 G6) 2612, 86x ot Oe 0 (3.6b) Strain-Displacement and Compatibility Relations: Only three strain-displacement relations are involved in most problems of fracture mechanics dealing with plates. If w, and u, are displacement components, strain components &,1, &) and £1, are expressed by well known relations, as: en= 4 (3.7a) Ixy auy f= (3.7) 2 (3.76) eqg= 2] 2 Ot 8.76) 2L ax, ax, Itis to be noted here that the tensorial strain has been chosen, whose normal components are the same as the normal components of the engineering strain, but a shear component is half the corresponding shear component of the engineering strain. The compatibility conditions (necessary relation between strain components) are obtained by eliminating u, and 1, from the above equations. Differentiating the equations, we have Fey _ Puy xy ax,axy Fey Puy ax” axfdx, Pe Puy Py Axdx, — 2[ axzox, ax,axzThe McGraw-Hill Companies Stress Intensity Factor 43 By substituting the first two equations in the last equation, and rearranging the terms, we obtain ae, Pep un , 12 2 +okn_p 2 fn 9 3.8) ax3 ax? Ax,ax (38) This relation between the strain components ensures compatibility. Stress-Strain Relations: For the linear isotropic materials deforming elastically, the stress-strain relations are well known as = aon =v (5x2 +633) ] (39a) f2> Hon -v(on +e)] 9) 1 fa> Flom -¥(n +o2)] 8.9) where E is the Young's Modulus, 1 is the shear modulus and vis the Poisson's Ratio. Plane Deformation: Consider a thin plate that is deformed in plane stress. On the free surfaces, the out of plane stresses are zero and they are usually negligible in the interior points of the plate. We thus assume O13 = O23 = O33 = 0 Therefore, the plate carries only in-plane stresses. ‘The stress-strain relations are simplified to &:= Hon von] (10a) 1 f= Glen -von] (3.10b) 2 _ 4¥) g (3.10¢) Qu E On the other hand, the plane strain case corresponds to a sufficiently thick plate for which (i) (a displacement in x; direction is restricted (w3=0) and (ii) variation in x; direction is zero lan = 0] ax, These two conditions yield £13 = &3 = &3 = 0 Therefore, for plane strain cases, simplified stress-strain equations are obtained by setting &3 = 0 in Eq, (3.9c) to have O33 = V (O11 + G22)el me Re ed 44 Elements of Fracture Mechanics Substituting 0,5 in Eqs (3.9a) and (3.96) and rearranging the terms, we have a-o (Ga) g= 3.11b) Also, Eq. (3.94) is manipulated to (1-v) (1-v’ 1-v? v_) coe 98 oye (art) on @ttg In the above equations, we define 11d) Ale) (3.12) We thus note that the form of these equations is exactly the same as that of the corresponding equations for plane stress cases [Fqs (3.10)] Biharmonic Differential Equation: When we glance at the field equations, we find that they are too many of them to compute. Therefore, we shall reduce the number of equations by going to the higher order of differential equations. The analysis of determining stress field in the vicinity of a crack tip can be done either in stress, strain or displacement components. Based on experience we find that it is convenient to solve the differential equations with stress components as dependent variables for many problems of fracture mechanics. This is because the boundary conditions, especially at the cracked faces, are usually known in stress components. In fact, the cracked surfaces are generally traction free, thereby making several stress components zero (¢.g., 63 = 0:2 = 0). Since we are planning to develop the differential equations in terms of stress components, compatibility conditions should be invoked. Thus, the compatibility condition of Eq. (3.8), which relates strain components should be changed into a relation between stress components. This is done by substituting stress-strain relations [Eq. (3.10)] into the compatibility relation [Eq- (3.8)] Thus, we have a e Fo, ae (Gy ~ VOn2) + ae (G22 — VOy,) - 2(1 + v) ede, 0 (3.13)