Fact-checked by Grok 2 weeks ago

Stress intensity factor

The stress intensity factor (SIF), denoted as K, is a fundamental parameter in linear elastic fracture mechanics (LEFM) that quantifies the intensity of the singular stress and displacement fields surrounding the tip of a in a brittle or quasi-brittle material subjected to remote loading or residual stresses. It provides a measure of the crack-tip stress state by incorporating the effects of applied load, crack length, and geometry, enabling the prediction of crack propagation and fracture behavior. The SIF was developed by George R. Irwin in 1957 as an extension of A.A. Griffith's 1920 energy-balance theory for brittle fracture, adapting it to account for plastic deformation in more ductile materials and establishing a practical framework for engineering analysis. Crack propagation in materials can occur under three primary modes of loading, each associated with a distinct : Mode I (K_I), the tensile opening mode where the crack faces separate to the crack plane; Mode II (K_{II}), the in-plane mode where the crack faces slide over one another in a direction to both the crack plane and the crack front; and Mode III (K_{III}), the out-of-plane (or anti-plane ) mode where the crack faces slide to the crack plane but to the crack front. These modes describe the relative displacement of the crack surfaces near the tip, with Mode I being the most common and critical in structural applications involving tensile stresses, as it typically requires the lowest critical load for propagation. In mixed-mode conditions, the effective SIF is often determined by algebraic of the individual mode components, reflecting the combined of multiple loading types. The magnitude of the SIF depends on factors such as the applied far-field \sigma, crack a, and component ; for an idealized through-crack of $2a in an infinite plate under uniform uniaxial , the Mode I SIF is expressed as K_I = \sigma \sqrt{\pi a}. More complex geometries require correction factors, such as K_I = 1.12 \sigma \sqrt{\pi a} for a single edge crack. Fracture initiates when the SIF reaches a material-specific critical value K_c (or K_{Ic} for plane strain Mode I conditions), termed the fracture toughness, which serves as a key measure of a material's resistance to crack growth and is determined experimentally under standardized conditions. This parameter has broad applications in aerospace, nuclear, and civil engineering for assessing structural integrity and preventing catastrophic failures.

Fundamentals

Definition and Physical Interpretation

The stress intensity factor (SIF), denoted as K, was introduced by George R. Irwin in 1957 as an extension of A. A. Griffith's energy-based criterion for brittle , specifically to characterize the local fields near the of cracks in materials. This emerged from analyses of distributions in cracked plates under , addressing the limitations of earlier energy approaches by focusing on the singular at the crack . Physically, the quantifies the amplitude of the near-tip singularity, serving as a scalar measure of how severely tip es are intensified by remote loading. It establishes a direct connection between macroscopic factors—such as applied , crack length, and structural —and the microscopic state immediately ahead of , enabling predictions of crack and without resolving the full . Notably, [K](/page/K) remains invariant under local coordinate transformations near the tip, making it a robust, geometry-independent descriptor of crack-tip loading severity. The validity of the relies on the foundational assumptions of linear fracture mechanics (LEFM), which posit that the material behaves in a linear , isotropic, and homogeneous manner throughout most of the body. Additionally, LEFM assumes small-scale yielding, wherein any deformation is confined to a tiny region at the crack tip, much smaller than the crack length or overall specimen size, ensuring that solutions dominate the stress analysis. Conceptually, the takes the general form K = \sigma \sqrt{\pi a} \, Y, where \sigma represents the nominal far-field , a is the relevant dimension (such as half-length for a central ), and Y is a dimensionless factor that incorporates the influence of the body's and loading configuration. This expression highlights how K scales with the of size, emphasizing the growing threat of longer cracks under fixed loading.

Modes of Loading

In , cracks can experience three fundamental modes of loading, each characterized by distinct crack face displacements and corresponding stress factors that quantify the stress state near the crack tip. These modes provide a framework for analyzing crack propagation under various loading conditions. Mode I (Opening Mode) represents the tensile loading where the crack faces separate perpendicular to the crack plane due to normal stresses acting normal to the crack surface. This mode is the most common and critical for brittle , as it directly promotes crack opening and extension. The stress factor K_I governs the intensity of the normal stresses near the crack tip in this mode, serving as a key parameter for predicting . Schematic diagrams for Mode I typically depict a straight crack with symmetric displacement vectors pointing away from the crack faces, illustrating the pure tensile separation. Mode II (In-Plane Shear or Sliding Mode) involves shear stresses parallel to the crack plane but perpendicular to the crack front, causing the crack faces to slide over each other in the plane of the crack. This mode arises in situations like anti-plane shear loading in configurations and is relevant for predicting sliding failures. The stress intensity factor K_{II} characterizes the in-plane intensity at the crack tip. Visual representations often show asymmetric displacement arrows along the crack faces, indicating relative sliding without out-of-plane motion. Mode III (Anti-Plane Shear or Tearing Mode) occurs under shear stresses parallel to both the crack plane and the crack front, leading to out-of-plane displacements where the crack faces tear relative to each other. This mode is less frequent but important in torsional or anti-plane loading scenarios, such as in shafts or plates under twisting. The stress intensity factor K_{III} quantifies the anti-plane intensity. Schematics for Mode III illustrate displacements perpendicular to the crack plane, resembling a tearing action with vectors pointing in opposite directions above and below the crack. In practice, real-world loading often results in mixed-mode conditions, where two or more stress intensity factors (e.g., K_I and K_{II}) are simultaneously non-zero, leading to complex crack paths that may kink or curve to align with the principal stress direction. These combinations complicate propagation predictions but are analyzed using criteria that consider the interaction of modes.

Theoretical Framework

Near-Tip Stress and Displacement Fields

The near-tip fields in linear elastic are analyzed using a centered at the crack tip, with the radial distance r measured from the tip and the angle \theta measured from (positive counterclockwise above the plane). This is valid in the asymptotic region where r is much smaller than the crack length or other characteristic dimensions of the body, capturing the dominant singular behavior. The stresses near the crack tip admit an asymptotic series expansion of the form \sigma_{ij} = \frac{K}{\sqrt{2\pi r}} f_{ij}(\theta) + \text{regular terms}, where K is the appropriate mode-specific , and the angular functions f_{ij}(\theta) are determined from the eigenvalue solution to the biharmonic equation, originally derived by Williams in 1957. This expansion separates the singular amplitude (controlled by K) from the angular variation, with higher-order terms becoming negligible close to the tip. The three fundamental modes of loading—Mode I (opening), Mode II (in-plane sliding), and Mode III (out-of-plane tearing)—each produce orthogonal fields, allowing mixed-mode problems to be superposed linearly. For Mode I loading, the in-plane stress components in Cartesian coordinates (with x along the crack plane ahead of the tip and y normal to it) are given by \begin{align} \sigma_{xx} &= \frac{K_I}{\sqrt{2\pi r}} \cos\frac{\theta}{2} \left[1 - \sin\frac{\theta}{2} \sin\frac{3\theta}{2}\right], \\ \sigma_{yy} &= \frac{K_I}{\sqrt{2\pi r}} \cos\frac{\theta}{2} \left[1 + \sin\frac{\theta}{2} \sin\frac{3\theta}{2}\right], \\ \sigma_{xy} &= \frac{K_I}{\sqrt{2\pi r}} \sin\frac{\theta}{2} \cos\frac{\theta}{2} \cos\frac{3\theta}{2}, \end{align} with \sigma_{zz} = \nu (\sigma_{xx} + \sigma_{yy}) under plane strain conditions (\nu is ) or \sigma_{zz} = 0 under ; the out-of-plane shear stresses \sigma_{xz} = \sigma_{yz} = 0. These expressions reflect the symmetric nature of Mode I, with maximum tensile along the crack plane (\theta = 0) where \sigma_{yy} \approx K_I / \sqrt{2\pi r}. The corresponding Mode I displacement fields, derived from integrating the strain-displacement relations with the stress field, are \begin{align} u_x &= \frac{K_I}{2\mu} \sqrt{\frac{r}{2\pi}} \cos\frac{\theta}{2} \left[ \kappa - 1 + 2 \sin^2 \frac{\theta}{2} \right], \\ u_y &= \frac{K_I}{2\mu} \sqrt{\frac{r}{2\pi}} \sin\frac{\theta}{2} \left[ \kappa + 1 - 2 \cos^2 \frac{\theta}{2} \right], \end{align} where \mu is the , and \kappa = 3 - 4\nu for plane strain or \kappa = (3 - \nu)/(1 + \nu) for . Along the crack faces (\theta = \pm \pi), these reduce to the crack face displacement u_y \approx (4 K_I / E) \sqrt{r / 2\pi} for (E is ), so the full crack opening displacement \delta = 2 u_y \approx (8 K_I / E) \sqrt{r / 2\pi}, highlighting the square-root dependence on distance behind the tip. For Mode II, the stress components exhibit antisymmetric behavior: \begin{align} \sigma_{xx} &= -\frac{K_{II}}{\sqrt{2\pi r}} \sin\frac{\theta}{2} \left[2 + \cos\frac{\theta}{2} \cos\frac{3\theta}{2}\right], \\ \sigma_{yy} &= \frac{K_{II}}{\sqrt{2\pi r}} \sin\frac{\theta}{2} \cos\frac{\theta}{2} \cos\frac{3\theta}{2}, \\ \sigma_{xy} &= \frac{K_{II}}{\sqrt{2\pi r}} \cos\frac{\theta}{2} \left[1 - \sin\frac{\theta}{2} \sin\frac{3\theta}{2}\right], \end{align} with displacements \begin{align} u_x &= -\frac{K_{II}}{2\mu} \sqrt{\frac{r}{2\pi}} \sin\frac{\theta}{2} \left[ \kappa + 1 - 2 \cos^2 \frac{\theta}{2} \right], \\ u_y &= \frac{K_{II}}{2\mu} \sqrt{\frac{r}{2\pi}} \cos\frac{\theta}{2} \left[ \kappa - 1 + 2 \sin^2 \frac{\theta}{2} \right]. \end{align} Mode II features shear-dominated stresses, with peak \sigma_{xy} along \theta = 0. For Mode III (antiplane ), the field simplifies to \tau_{yz} = -\frac{K_{III}}{\sqrt{2\pi r}} \sin\frac{\theta}{2}, \tau_{xz} = \frac{K_{III}}{\sqrt{2\pi r}} \cos\frac{\theta}{2}, and displacement w = \frac{2 K_{III}}{\mu} \sqrt{\frac{r}{2\pi}} \sin\frac{\theta}{2} (out-of-plane), from in-plane modes. The characteristic $1/\sqrt{r} singularity in the leading-order stress terms implies theoretically infinite stresses at r = 0, an idealization that holds for continuum linear elasticity but breaks down at atomic scales where nonlinear effects dominate; this singularity underscores the stress intensity factor's role in quantifying crack-tip severity.

Relation to Energy Release Rate

The energy release rate G quantifies the driving force for crack propagation in linear elastic fracture mechanics (LEFM), defined as the decrease in the total potential energy per unit crack extension per unit thickness, G = -\frac{1}{B} \frac{d\Pi}{da}, where \Pi is the potential energy, a is the crack length, and B is the specimen thickness. This concept originates from Griffith's analysis of brittle fracture, where unstable crack growth occurs when G equals the critical value G_c = 2\gamma, with \gamma representing the surface energy required to create new crack surfaces. Irwin extended Griffith's energy balance by linking G directly to the stress intensity factor K, establishing the fundamental relation G = \frac{K_I^2}{E'}, where E' = E for and E' = \frac{E}{1 - \nu^2} for plane strain, with E as and \nu as . This equivalence demonstrates that K, which characterizes the local stress singularity at the crack tip, is thermodynamically connected to the global energy perspective provided by G. For combined in-plane modes, the relation generalizes to G = \frac{K_I^2 + K_{II}^2}{E'}, while including out-of-plane shear yields G = \frac{K_I^2 + K_{II}^2}{E'} + \frac{(1 + \nu) K_{III}^2}{E}, where the mode III term arises from antiplane shear contributions. The connection between G and K can be derived using the compliance method, where G = \frac{P^2}{2B} \frac{dC}{da} and C = \frac{\delta}{P} is the (with \delta as load-point and P as applied load), or through crack extension, which computes the from near-tip and fields. (Note: Specific URL for ASTM proceedings 1954; assuming access via https://www.astm.org) In both approaches, substituting the asymptotic near-tip fields proportional to K yields the G\propto K^2 scaling, confirming that K encapsulates the intensity of energy available for growth under LEFM assumptions of small-scale yielding and . This relation holds equivalently under or plane strain conditions, with the adjustment in E' accounting for effects on the field. Physically, while K offers a local measure of stress amplification near the crack tip, G provides a global view of the energy balance driving , bridging microscopic effects with macroscopic ; their equivalence underscores the robustness of LEFM for predicting crack stability across scales.

Connection to Path-Independent Integrals

The serves as a path-independent contour that characterizes the of the crack-tip fields in , applicable to both linear and nonlinear elastic materials. It is defined for a two-dimensional field as J = \oint_{\Gamma} \left( W \, dy - \mathbf{T} \cdot \frac{\partial \mathbf{u}}{\partial x} \, ds \right), where \Gamma is a closed contour encircling the crack tip (traversed counterclockwise), W is the strain energy density, \mathbf{T} is the traction vector on the contour, \mathbf{u} is the displacement vector, and ds is the differential arc length along \Gamma. This formulation captures the energy flow toward the crack tip without relying on the singular fields directly. The path independence of the arises from its mathematical structure, proven using the under conditions of , absence of body forces, and no singularities within the region enclosed by the contour. Specifically, the difference between J evaluated on two arbitrary contours \Gamma_1 and \Gamma_2 surrounding integrates to zero over the annular region between them, as the integrand's divergence vanishes in . This property holds for hyperelastic materials, including nonlinear -plastic behaviors, provided the material response is derivable from a potential, making J a robust measure even beyond strict linear (LEFM). In LEFM, the J-integral equates to the energy release rate G, such that J = G, and for mode I loading under plane strain conditions, this yields J = G = \frac{K_I^2}{E'}, where K_I is the mode I stress intensity factor and E' = \frac{E}{1 - \nu^2} is the effective modulus with E and \nu. Extensions to mixed-mode loading incorporate contributions from modes II and III via vectorial or tensorial forms of J, maintaining the equivalence J = G while linking to the full stress intensity factor vector \mathbf{K}. This connection was established by James R. Rice in , who introduced the as a generalization of George Irwin's energy-based approaches, particularly for cases of non-proportional loading where traditional stress intensity factors alone are insufficient. The path independence of J enables its evaluation on contours far from the crack-tip singularity, facilitating practical computations in finite element analysis and experimental setups where mesh refinement near the tip is challenging or inaccurate. This utility has made J a cornerstone for numerical of processes, allowing reliable estimation of crack-tip intensity parameters without direct resolution of asymptotic fields.

Fracture Toughness

Critical Stress Intensity Factor

The critical stress intensity factor, often denoted as K_c, is a fundamental material property known as , representing the threshold value of the stress intensity factor at which a initiates unstable under monotonic loading conditions. In linear elastic (LEFM), growth occurs when the applied stress intensity factor K reaches or exceeds K_c, providing a to predict brittle in materials with preexisting cracks. For mode I tensile loading under plane-strain conditions, the critical value is specifically K_{Ic}, a constant that quantifies a material's resistance to crack extension in the thickness direction where out-of-plane strains are constrained. Similar critical values, K_{IIc} and K_{IIIc}, apply to modes II and III, respectively. Typical K_{Ic} values span a wide range depending on class; brittle exhibit low toughness around 0.7 MPa\sqrt{\text{m}}, while ductile metals like alloy steels can achieve values exceeding 100 MPa\sqrt{\text{m}}.
MaterialK_{Ic} (MPa\sqrt{\text{m}})
Soda-lime glass0.7
(quenched and tempered)50–60
150–200
The magnitude of K_c is sensitive to several environmental and geometric factors, including , specimen thickness, and loading . Lower temperatures generally reduce K_c by promoting brittle behavior. The effect of on K_c is material-dependent; in some cases, higher s can decrease crack initiation toughness due to limited time for plastic relaxation. In thinner specimens, plane-stress conditions allow greater , yielding a higher apparent K_c compared to the more conservative plane-strain K_{Ic} in thick sections. For ductile materials, K_c may not remain constant but instead follows an R-curve, where crack resistance increases with crack extension due to mechanisms like crack bridging or deflection. Standardized testing ensures reliable measurement of K_{Ic}, with ASTM E399 specifying procedures for metallic materials using fatigue-precracked specimens under three-point bending or tensile loading. This standard mandates valid conditions such as small-scale yielding, where the plastic zone remains negligible relative to crack and specimen dimensions, and minimum thickness B \geq 2.5 (K_{Ic}/\sigma_{ys})^2 to enforce plane strain. Historically, measured K_{Ic} values for brittle materials closely align with predictions from A.A. Griffith's 1920 theory of fracture, which linked critical stress to release for crack propagation in ideal brittle solids like . G.R. Irwin's 1957 development of the stress intensity factor extended Griffith's energy-based ideas into a stress-field approach, enabling practical quantification of across materials.

Energy-Based Criteria

The energy-based criteria for fracture initiation and propagation in brittle materials originated with the Griffith criterion, which posits that unstable crack growth occurs when the energy release rate G equals twice the surface energy per unit area \gamma, expressed as G = 2\gamma. This threshold represents the balance between the elastic energy released by crack advance and the energy required to create new crack surfaces, providing a thermodynamic foundation for brittle fracture under tensile loading. In its original formulation, the criterion applies to ideally brittle solids like glass, where no significant plastic deformation occurs, and the surface energy \gamma is determined from thermodynamic considerations of atomic bond breaking. Subsequent developments introduced the critical energy release rate G_c as a material property analogous to the critical stress intensity factor K_c, with the equivalence given by G_c = \frac{K_c^2}{E'} for plane strain conditions, where E' = \frac{E}{1 - \nu^2}, E is the , and \nu is . This relation links energy-based and stress-based approaches within linear elastic (LEFM), allowing G_c to characterize fracture resistance in scenarios involving or viscoelastic materials, where rate effects influence energy dissipation beyond simple surface creation. For instance, in dynamic , G_c accounts for contributions during rapid crack propagation, while in viscoelastic polymers, it incorporates time-dependent relaxation that alters the effective energy threshold. To extend the Griffith criterion to materials exhibiting limited plasticity, such as metals, the Dugdale model incorporates plastic zones ahead of the crack tip by assuming a strip-yield region where yielding occurs at the finite yield stress, effectively shielding the crack tip singularity and adjusting the energy release rate to include plastic work. This cohesive zone approach predicts the size of the plastic zone and modifies the effective G to prevent infinite stresses, providing a more realistic estimate of fracture load for ductile-brittle transitions. Similarly, the Orowan modification augments the surface energy term \gamma with a plastic dissipation component \gamma_p, yielding G = 2(\gamma + \gamma_p), to capture energy absorbed through localized yielding near the crack surfaces without assuming large-scale plasticity. This adjustment reconciles the criterion with experimental observations in quasi-brittle metals, where small plastic zones contribute significantly to toughness. The G-criterion offers advantages over the K_c-based approach for rate-dependent materials, such as composites or polymers, because it directly quantifies availability and dissipation, facilitating analysis of time-varying loading where stress fields alone may not suffice; however, it converges to the K_c limit under the small-scale yielding assumptions of LEFM. Despite these strengths, the criterion assumes quasi-static conditions and negligible inertia effects, limiting its applicability to high-speed fractures, and it breaks down in cases of extensive where global yielding dominates energy balance.

Analytical Solutions for Common Geometries

Central Crack in Infinite Plate

The central in an infinite plate serves as a fundamental benchmark geometry in linear elastic , providing an exact analytical solution for intensity factors under remote loading. This configuration involves an infinite, isotropic, linear elastic plate containing a straight through-thickness of total length $2a, with the oriented perpendicular to a uniform uniaxial tensile \sigma applied at infinity, resulting in Mode I opening. The solution assumes or plane strain conditions, negligible body forces, and small-scale yielding where the tip fields dominate locally. The Mode I stress intensity factor for this geometry is K_I = \sigma \sqrt{\pi a}, where a is the half-crack length. This expression emerges from the Westergaard stress function approach, which employs complex variable methods to derive the Airy stress function for the entire stress field in the plate: Z(z) = \frac{\sigma}{\sqrt{1 - (a/z)^2}}, where z = x + iy is the complex coordinate. Near the crack tip at z = a, the stresses exhibit the characteristic $1/\sqrt{r} singularity, matching the asymptotic near-tip fields \sigma_{ij} \approx \frac{K_I}{\sqrt{2\pi r}} f_{ij}(\theta), with r and \theta as polar coordinates from the tip; far from the crack, the stresses recover the uniform remote field \sigma_{yy} = \sigma. The stress intensity factor concept was formalized by Irwin from this singular field behavior. An equivalent derivation uses conformal mapping, transforming the cracked plate to an uncracked domain via the mapping z = \frac{w + \sqrt{w^2 - a^2}}{2}, yielding the same K_I. Extensions to biaxial remote loading, with principal stresses \sigma_y (perpendicular to the crack) and shear \sigma_{xy}, yield K_I = \sigma_y \sqrt{\pi a} and K_{II} = -\sigma_{xy} \sqrt{\pi a}, maintaining the infinite plate assumptions and linear elasticity. These factors characterize mixed-mode conditions while preserving the uniform far-field stress state.

Edge Crack in Semi-Infinite Plate

The edge crack in a semi-infinite plate represents a fundamental configuration in linear elastic fracture mechanics, modeling a straight through-thickness crack of length a perpendicular to the free edge of an elastic half-plane under uniform uniaxial tensile loading \sigma applied parallel to the edge at infinity. The plate occupies the region x \geq 0, with the crack extending from the origin along the negative x-axis to (-a, 0), the crack faces remaining traction-free, and the plate edge (x=0) also traction-free. This setup captures the influence of a nearby free boundary on the crack-tip stress field, distinguishing it from unbounded geometries. The mode I stress intensity factor K_I for this problem is given by K_I = 1.122 \, \sigma \sqrt{\pi a}, where the geometry correction factor of 1.122 arises from the boundary effects that amplify the stress concentration at the crack tip compared to an infinite domain. This exact solution is derived using the complex variable method of Muskhelishvili, involving conformal mapping to transform the domain and solve the resulting Hilbert problem for the stress analytic function. In contrast to the central crack in an infinite plate, where K_I = \sigma \sqrt{\pi a} for a of full $2a, the edge crack experiences an approximately 12% higher K_I due to the free edge singularity, which reduces the effective constraint on crack opening displacements. This amplification highlights the role of geometric boundaries in elevating risk. This is widely applied to approximate surface-breaking flaws in large-scale components, such as edge notches in pressure vessels or fatigue cracks initiating from machined surfaces in structures, where the crack length is small relative to the overall dimensions.

Embedded Circular Crack

The embedded circular crack, commonly termed the penny-shaped crack, represents a canonical three-dimensional fracture geometry in linear elastic fracture mechanics. It features a flat, circular crack of radius a lying in a plane within an infinite isotropic elastic medium, loaded by a uniform remote tensile stress \sigma directed normal to the crack plane. This setup induces primarily mode I fracture, characterized by crack face separation perpendicular to the plane, with the crack faces remaining traction-free within the circular region and the surrounding material maintaining continuity of displacement and traction. The mode I stress intensity factor K_I for this configuration is constant along the entire circular crack front due to the axisymmetric geometry and loading. Its value is given by K_I = 2 \sigma \sqrt{\frac{a}{\pi}}. This yields an average K_I of $2 \sigma \sqrt{a/\pi}, which is reduced from the two-dimensional central crack counterpart \sigma \sqrt{\pi a} by the factor $2/\pi \approx 0.637, accounting for three-dimensional stress redistribution around the crack periphery. The analytical solution originates from Sneddon's application of integral transforms, specifically Hankel transforms of order zero, to resolve the mixed boundary conditions of the axisymmetric elasticity problem in cylindrical coordinates. This approach yields closed-form expressions for the stress and displacement fields, from which K_I is extracted as the coefficient of the 1/\sqrt{2\pi \rho}} singularity ahead of the front, where \rho is the in-plane distance from the tip. An equivalent derivation employs the analogy to a thin ellipsoidal inclusion under uniform stress, taking the oblate spheroid limit as the minor axis approaches zero, as established by Eshelby's inclusion theory. Although the intensity factor itself remains uniform along the crack front in this symmetric case, three-dimensional effects manifest in the overall field, with the crack opening displacement peaking at the center (r=0) and tapering to zero at the edges (r=a), influencing local concentrations. Mode III (tearing) is negligible under pure tensile loading, as the antisymmetric response is not excited by the uniform, perpendicular field.

Slanted Crack in Biaxial Field

In an infinite plate subjected to remote biaxial principal stresses \sigma_x and \sigma_y, consider a central through-crack of $2a inclined at an \beta to axis aligned with \sigma_x. The geometry represents a classic case for analyzing mixed-mode fracture under uniform far-field loading, where faces remain traction-free.90049-8) The stress intensity factors at the crack tips are obtained through a coordinate transformation that resolves the remote stresses into components normal and parallel to the crack plane. Specifically, K_I = (\sigma_x \cos^2 \beta + \sigma_y \sin^2 \beta) \sqrt{\pi a} K_{II} = (\sigma_y - \sigma_x) \sin \beta \cos \beta \sqrt{\pi a} These expressions capture the mode I opening due to the normal stress component and the mode II sliding due to the resolved shear stress. The derivation relies on rotating the stress tensor to the crack-oriented coordinates and applying the known solution for a perpendicular central crack in an infinite plate.90049-8) When \beta = 0^\circ, the formulas reduce to K_I = \sigma_x \sqrt{\pi a} and K_{II} = 0, corresponding to pure mode I loading aligned with the central crack case in an infinite plate under uniaxial tension \sigma_x. At \beta = 90^\circ, K_I = \sigma_y \sqrt{\pi a} and K_{II} = 0, yielding pure mode I under \sigma_y. For intermediate angles, such as \beta = 45^\circ under uniaxial \sigma_x (with \sigma_y = 0), both K_I and K_{II} are nonzero, with magnitudes K_I = K_{II} = \frac{1}{2} \sigma_x \sqrt{\pi a} (up to sign convention for K_{II}).90049-8) The presence of nonzero K_{II} induces mixed-mode conditions, where crack propagation typically involves kinking from the initial plane to align with the direction of maximum tangential , effectively driven by the component to eliminate mode II at the new tip. This behavior is predicted by criteria such as the maximum tangential , which determines the kink angle \theta_k \approx -2 \arctan \left( \frac{K_{II}}{K_I} \right) for small s. The combined modes amplify the local intensity near the tip compared to pure mode I, as the effective intensity—often assessed via \sqrt{K_I^2 + k K_{II}^2} where k depends on the fracture criterion—exceeds that of either mode alone, influencing critical loads and failure paths in assessments.90049-8)

Experimental and Specimen Configurations

Compact Tension Specimen

The compact tension (CT) specimen is a standardized notched rectangular used primarily for mode I testing in metallic materials. It features a central through-thickness of a in a rectangular plate of width W and thickness B, with pin-loaded arms for tensile loading along the crack plane. The standard proportions, as specified by ASTM E399, include a of approximately $1.25W and pin hole diameters of $0.25W, ensuring consistent distribution and ease of precracking. The stress intensity factor K_I for the CT specimen is calculated using the formula: K_I = \frac{P}{B \sqrt{W}} (2 + \alpha) \frac{0.886 + 4.64\alpha - 13.32\alpha^2 + 14.72\alpha^3 - 5.6\alpha^4}{(1 - \alpha)^{3/2}} where \alpha = a/W is the normalized crack length, P is the applied load, B is the specimen thickness, and W is the specimen width. This polynomial expression provides the geometry correction factor for the nominal stress intensity. The formula is valid for $0.45 \leq \alpha \leq 0.55, a range that balances stability during testing and minimizes , with accuracy derived from finite calibrations against experimental . Outside this range, alternative corrections or specimen geometries may be required to maintain precision. The specimen's compact size (typically W \approx 50 mm) and simple pin-loading mechanism facilitate testing in standard laboratory setups, making it suitable for determining the critical stress intensity factor K_{Ic} in materials ranging from brittle metals to those with high toughness. To ensure plane-strain conditions and accurate crack propagation, corrections are applied for side grooving, which increases by reducing lateral deformation, and for precracks, typically 0.15 mm to 0.25 mm long to simulate sharp natural flaws without introducing . These modifications are integral to ASTM E399 protocols for valid K_{Ic} measurements.

Single-Edge Notched Bend Specimen

The single-edge notched bend (SENB) specimen consists of a rectangular with width W, thickness B, and support span S = 4W, featuring an edge of length a machined from one long side and subjected to three-point under load P applied at the 's above the . This generates a mode I stress field suitable for evaluation. The mode I stress intensity factor K_I for the SENB specimen is expressed as K_I = \frac{P S}{B W^{3/2}} f(\alpha), where \alpha = a/W is the normalized crack length, and the geometry correction function f(\alpha) is f(\alpha) = \frac{3 \alpha^{1/2} \left[1.99 - \alpha (1 - \alpha) (2.15 - 3.93 \alpha + 2.7 \alpha^2)\right] }{2 (1 + 2\alpha) (1 - \alpha)^{3/2}}. This , derived from boundary solutions to the stress function, provides accuracy within 0.5% over $0 < \alpha < 1 and forms the basis for standard calculations in testing. ASTM E399 specifies this expression for f(\alpha) applicable in the range $0.45 \leq \alpha \leq 0.55 for metallic materials. Under ASTM E399, the SENB geometry is used for metallic materials to promote stable crack propagation and reliable measurement of plane-strain K_{Ic} under controlled conditions; for brittle materials like ceramics, other standards such as ASTM C1421 apply. The formulation assumes linear elastic conditions, with valid results requiring a , fatigue-precracked front perpendicular to the specimen faces to ensure uniform stress intensity across the thickness.

Other Standard Test Geometries

The double (DCB) specimen is a widely adopted for evaluating mode I interlaminar in materials, particularly for resistance. In this setup, two arms are separated by a load P applied at the unloaded end, with crack length a, specimen width b, and arm height h/2 (total height h). Using classical theory, the mode I stress intensity factor is approximated as K_I = \frac{12 P a}{b h^{3/2}} This expression assumes Euler-Bernoulli beam behavior and neglects root rotation effects, providing a simple yet effective means for fracture characterization in unidirectional laminates. The DCB test is standardized under ASTM D5528 and ISO 15024, which primarily focus on determining the critical energy release rate G_{Ic} from load-displacement data, with K_I derived via the relation G_{Ic} = K_{Ic}^2 (1 - \nu^2)/E for plane strain conditions. For mode II shear loading, the end-notched flexure (ENF) specimen applies three-point bending to a beam with an initial delamination of length a, span L, width b, and height h, promoting sliding shear along the crack plane. The mode II stress intensity factor, based on Timoshenko beam theory accounting for shear deformation and root rotation, is given by K_{II} = \frac{3 P L}{2 b h^{3/2}} \sqrt{\frac{a}{L^2 + (11/10) a^2}} This formula ensures accurate representation of the stress field near the crack tip, particularly for composite interlaminar where II dominance is critical for predicting -induced . The ENF configuration is outlined in ASTM D7905/D7905M and ISO 15114, enabling measurement of the critical II release rate G_{IIC} under controlled conditions. Disk-shaped specimens, such as the centrally cracked disk (CCBD), facilitate mixed-mode I/II fracture testing by subjecting a circular disk of D and thickness b to diametral P, with a central of length 2a inclined at β to the loading . The mixed-mode intensity factors are expressed as K_I = f_I(α, β) P / \sqrt{π a} and K_{II} = f_{II}(α, β) P / \sqrt{π a}, where α = a/D and the functions f_I, f_{II} are derived from analytical field solutions or finite element calibration for wide parameter ranges. These tests are valuable for rocks, , and brittle materials, revealing envelopes under combined tension-shear loading. Chevron-notched variants, like the cracked chevron-notched disk (CCNBD), enhance in mixed modes by introducing a V-shaped that constrains growth, with K solutions following similar normalized forms K = Y(α) P / \sqrt{π a_0} where Y incorporates and loading for mode I/II partitioning. Such configurations are standardized in ASTM E1304 for ceramics and rocks, supporting comprehensive mixed-mode analysis. For composites and adhesives, ASTM D6671 provides protocols for mixed-mode I/II interlaminar toughness using modified geometries, addressing under arbitrary mode ratios up to recent validations in 2022.

Computational Methods

Finite Element Approaches

Finite element methods provide a versatile numerical approach for computing intensity factors in geometries too complex for analytical solutions, by discretizing the into and solving the linear elastic equations while specially treating the crack tip . These methods fall into direct approaches, which extrapolate singular or displacement fields from near-tip nodal values, and indirect approaches based on path-independent integrals like the , where the intensity factor relates to J via J = \frac{K_I^2 (1 - \nu^2)}{E} for plane strain or J = \frac{K_I^2}{E} for . Special crack-tip and mesh refinement ensure accurate capture of the $1/\sqrt{r} , enabling reliable predictions in engineering applications. Commercial finite element software such as and incorporate these approaches, offering built-in contour or integral evaluations for J and direct SIF extraction via quarter-point or enriched , with options for automatic mesh refinement near the crack tip to balance accuracy and efficiency. These tools support , , and dynamic analyses, often requiring collapsed at the tip for cracks. Validation studies confirm their reliability, with computed SIFs converging to within 1-2% of analytical values for benchmark geometries like the infinite plate with central crack under remote tension when using refined meshes (e.g., 10-20 along the crack). For complex or multi-crack bodies, finite element methods excel by handling arbitrary loading, material interfaces, and geometries without closed-form restrictions, though computational cost increases with refinement. Modern extensions include the (XFEM), which enriches standard elements with discontinuity functions to model crack propagation without remeshing, allowing efficient SIF computation in evolving crack scenarios. Direct methods rely on extrapolating stresses or displacements near the crack tip using modified finite elements that inherently reproduce the singular field. Quarter-point elements achieve this by shifting the midside nodes of standard quadratic isoparametric elements to the quarter-point location along edges connected to the crack tip, distorting the element shape to yield the desired $1/\sqrt{r} stress variation while maintaining isoparametric mapping. This technique, originally developed for triangular elements in and analysis, was extended to general elements and has become standard for two- and three-dimensional problems. Stresses or displacements from these elements are then extrapolated to the crack tip boundary (typically r approaching 0) to estimate the stress intensity factors. In displacement-based direct methods, the mode I stress intensity factor is obtained from near-tip nodal displacements via asymptotic field relations. For a quarter-point element, the standard formulation is K_I = E' v_g \sqrt{\frac{2\pi}{L}}, where v_g is the transverse of the quarter-point node on one face at distance L/4 from the (with L the element length), E' = E for or E/(1 - \nu^2) for plane strain. This expression derives from matching the computed displacements to the leading-order singular term in the Williams expansion, providing accurate K values with minimal s around the when combined with quarter-point modeling. Similar relations apply for mode II using in-plane shear displacements. Indirect methods employ the , computed numerically in finite element analysis to avoid direct singularity handling. The (or equivalent) integral form transforms the into a over a enclosing the tip, given by J = \int_A \left[ q \frac{\partial W}{\partial x} - W \frac{\partial q}{\partial x} + \mathbf{T} \cdot \frac{\partial (q \mathbf{u})}{\partial x} \right] dA, where A is the between two contours, W is the , \mathbf{u} and \mathbf{T} are displacements and tractions, x is the crack advance direction, and q is a smooth virtual crack extension function (unity inside the inner contour, zero outside the outer). This formulation, equivalent to the original contour but less sensitive to mesh distortion near the tip, facilitates automated computation and path independence verification. The derivative technique underpins early implementations, relating J to changes in system with crack extension. Stress intensity factors are then extracted from J using the relation above, with mixed-mode extensions for K_{II} and K_{III}.

Boundary Integral Techniques

The (BEM) provides an efficient framework for computing stress intensity factors (SIFs) in linear elastic fracture mechanics by solving boundary integral equations derived from the Kelvin fundamental solution for isotropic elastostatics. This approach discretizes only the problem boundaries, transforming three-dimensional volume problems into two-dimensional surface integrations, which significantly reduces complexity. Cracks are incorporated through specialized modeling techniques, such as double-node elements on crack surfaces to capture displacement discontinuities while enforcing traction-free boundary conditions on both faces. Seminal developments in this formulation, including multi-domain decompositions for complex geometries, were advanced in the late 1970s and , enabling accurate representation of embedded or surface cracks in infinite or semi-infinite domains. Extraction of SIFs in BEM relies on post-processing the solutions to evaluate near-tip fields. techniques include analyzing relative crack-face displacements or tractions, often via quarter-point elements that asymptotically match the square-root . A direct method utilizes the hypersingular traction , where the mode I SIF K_I emerges in the as the collocation point approaches the crack tip along the crack extension: K_I = \lim_{\rho \to 0^+} \frac{2\mu}{\kappa + 1} \sqrt{\frac{2\pi}{\rho}} \Delta u_y(\rho), with \mu as the shear modulus, \kappa depending on Poisson's ratio, \rho the distance from the tip, and \Delta u_y the crack opening displacement; similar expressions apply for modes II and III. Equivalent approaches employ the domain or path-independent adapted to boundary data or energy release rate computations from virtual crack extensions. These methods yield high accuracy for mixed-mode SIFs, particularly in two- and three-dimensional configurations. BEM excels in fracture problems by inherently incorporating the stress singularity through the fundamental solution, avoiding the need for volumetric meshing and special crack-tip enrichments required in domain-based methods. This boundary-only discretization is especially advantageous for infinite domains, multi-crack interactions, and problems with curved crack fronts, where it maintains precision with fewer degrees of freedom—typically achieving SIF errors below 1% for benchmark cases like the infinite plate with central crack. In contrast to standard finite element analysis, BEM provides superior accuracy for traction-free cracks without auxiliary elements, as the integral equations naturally resolve the $1/\sqrt{r} near-tip behavior. Commercial implementations like FRANC3D and BEASY leverage these features for three-dimensional SIF computations and crack growth simulations, supporting adaptive remeshing for evolving fronts and integration with finite element solvers for hybrid analyses.

Applications and Limitations

Engineering Design and Failure Prediction

In engineering design, the stress intensity factor (SIF) serves as a critical parameter for ensuring structural integrity under potential crack propagation, particularly in damage-tolerant approaches where components are evaluated to withstand loads even after flaw . In applications, (FAA) Advisory Circular 25.571 mandates damage tolerance evaluations that incorporate principles, including SIF calculations to predict crack growth and residual strength, requiring the structure to sustain ultimate loads (1.5 times limit loads) with detectable damage while ensuring the applied SIF K does not exceed the K_c at those load levels. Similarly, in , ASME Boiler and Code Section XI (IWB-3612) employs SIF in flaw evaluations for pressure vessels, comparing the computed SIF to K_c with safety factors of \sqrt{10} (≈3.16) for normal/upset conditions and \sqrt{2} (≈1.41) for emergency/faulted conditions to prevent brittle failure. For fatigue-prone structures, the SIF range \Delta K is integrated into crack growth models to forecast propagation rates and establish inspection intervals. The seminal Paris law describes stable crack growth as \frac{da}{dN} = C (\Delta K)^m, where da/dN is the crack extension per cycle, \Delta K = K_{\max} - K_{\min} is the SIF range, and C and m are empirical material constants derived from testing; this relation allows numerical integration from initial flaw size to a critical length, yielding cycles-to-failure estimates that inform non-destructive inspection schedules in high-cycle applications like aircraft wings. Probabilistic methods enhance SIF-based predictions by accounting for variability in flaw sizes and material properties, using distributions to model failure risks. Flaw detection thresholds often follow log-normal or extreme value distributions, while fracture toughness scatter is captured by Weibull models, such as the Beremin approach, which applies a two-parameter Weibull distribution to the local stress criterion \sigma_u for cleavage fracture, enabling Monte Carlo simulations to compute failure probabilities under distributed initial flaws and K_c values. Real-world applications underscore the consequences of SIF oversight, as seen in the 1988 Aloha Airlines Flight 243 incident, where undetected multiple-site cracks in the fuselage lap joints—propagating under cyclic pressurization—led to explosive decompression at 24,000 feet, with post-accident analysis revealing exceedance due to corrosion-assisted cracking beyond inspection thresholds. In contrast, modern fiber metal laminates (FMLs), such as used in the fuselage, leverage fiber bridging to reduce effective in metal layers, enhancing damage tolerance by slowing crack growth rates by up to 80% compared to monolithic aluminum under equivalent loading. Software tools like NASGRO, developed by NASA and Southwest Research Institute, integrate SIF solutions with the Paris law and closure effects to predict \Delta K-driven crack growth lives for complex geometries, supporting probabilistic life assessments in aging aircraft programs by simulating inspection intervals and residual strength margins.

Assumptions of Linear Elastic Fracture Mechanics

Linear elastic fracture mechanics (LEFM) relies on several fundamental assumptions to characterize crack-tip stress fields through the stress intensity factor (SIF). The material is assumed to behave linearly elastically everywhere except in a minuscule plastic zone at the crack tip, where yielding is confined to a region much smaller than the crack length (r_p << a). Additionally, the material is considered homogeneous and isotropic, ensuring uniform elastic properties without directional variations. Loading is quasi-static and monotonic, avoiding dynamic effects or large rigid-body rotations that could invalidate the elastic field solutions. For LEFM to yield valid SIF results, specific criteria must be met to enforce plane strain conditions and small-scale yielding. Specimen dimensions, including crack length (a), thickness (B), and remaining (W - a), must satisfy B, a, (W - a) ≥ 2.5 (K_c / σ_y)^2, where K_c is the and σ_y is the yield strength, ensuring the plastic zone remains negligible relative to structural features. This criterion, derived from standards for fracture toughness testing, confirms that the field is dominated by the singularity rather than . Monotonic loading without significant unloading or rotations further upholds the assumption of proportional stress-strain responses. LEFM's assumptions limit its applicability, particularly for ductile materials where extensive plasticity violates small-scale yielding. In such cases, elastic-plastic fracture mechanics (EPFM) employs parameters like the or crack-tip opening (CTOD) to account for larger plastic zones. For dynamic or high-strain-rate fracture, the quasi-static assumption fails, necessitating modified formulations that incorporate rate-dependent effects. Recent extensions beyond traditional LEFM address small-scale nonlinearity while preserving core elastic dominance. Models incorporating limited inelasticity near the crack tip extend validity to mildly ductile behaviors without fully resorting to EPFM. Phase-field methods, evolving since the , model diffuse cracks in heterogeneous materials by regularizing sharp discontinuities, enabling simulations of complex fracture paths in anisotropic composites up to 2025 applications. These approaches bridge LEFM gaps in nonlinear and multi-scale scenarios, such as soft materials or under varying rates.

References

  1. [1]
    [PDF] Introduction to Fracture Mechanics - MIT
    Jun 14, 2001 · As the temperature is lowered, the yield stress σY will increase as described in Module 20, and the fracture stress σf will decrease (since ...
  2. [2]
    2.16.2 Stress intensity factor extraction
    The stress intensity factors , , and play an important role in linear elastic fracture mechanics. They characterize the influence of load or deformation on ...
  3. [3]
    [PDF] 7. fracture
    fracture will occur when the stress intensity factor in the component is equal to the fracture toughness of the material such that: Fractures OCCURS if KIc ...
  4. [4]
    [PDF] fracture mechanics
    This critical value of the stress intensity factor is known as the fracture toughness of the material. The fracture toughness can be considered the limiting ...
  5. [5]
    [PDF] The Merging of Fatigue and Fracture Mechanics Concepts
    5--Dominant fatigue-crack-closure mechanisms. (after Suresh and Ritchie, 1982 [36]). STRESS-INTENSITY. FACTORS. The essential feature of fracture mechanics is ...
  6. [6]
    Analysis of Stresses and Strains Near the End of a Crack Traversing ...
    Jun 10, 2021 · It is shown that the other parameter, called the stress-intensity factor, is proportional to the square root of the force tending to cause crack extension.
  7. [7]
    Stress Intensity Factor - an overview | ScienceDirect Topics
    The stress intensity factor, K, is defined as a measure of the severity of a crack situation influenced by crack size, stress, and geometry, assuming linear ...
  8. [8]
    Stress Intensity Factor - Fracture Mechanics
    Introduction. The stress intensity factor was developed in 1957 by George R Irwin, the man usually considered to be the father of fracture mechanics [1].Missing: paper | Show results with:paper
  9. [9]
    Stress Intensity Factor | Engineering Library
    Firstly, the stress intensity factor defines the amplitude of the crack tip singularity, and consequently the intensity of the local stress field. Local ...
  10. [10]
    [PDF] Fracture Mechanics
    An important assumption of Linear Elastic Fracture Mechanics is that the plastic region around the crack itself is very small relative to the remainder of ...
  11. [11]
    Fracture Mechanics - MechaniCalc
    In fracture mechanics, a stress intensity factor is calculated as a function of applied stress, crack size, and part geometry. Failure occurs once the stress ...
  12. [12]
    Loading Modes I, II, III - Fracture Mechanics
    Mode I is the Opening Mode, Mode II is in-plane shear, and Mode III is the Tearing Mode, driven by out-of-plane shear stresses.
  13. [13]
    The science and engineering of fracture - Article
    Irwin, “Analysis of stresses and strains near the end of a crack traversing a plate,” J. Appl. Mech.,24 361–364 (1957). Google Scholar. T. Ishikawa ...<|control11|><|separator|>
  14. [14]
  15. [15]
    Determination of mixed-mode stress intensity factors, fracture ...
    A numerical and experimental investigation for determining mixed-mode stress intensity factors, fracture toughness, and crack turning angle for BX-265 foam ...
  16. [16]
    [PDF] The Elastic Stress Field around a Crack Tip - Elsevier
    The stress field around a crack tip governs brittle fracture. It's analyzed using methods like the complex potential function method, and the stress intensity ...<|control11|><|separator|>
  17. [17]
    [PDF] On the Stress Distribution at the Base of a Stationary Crack1 1) 8
    For the antisymmetrical stress distribution the distortion strain energy is a relative maximum along the crack and. 60 per cent lower ± 85 deg to the sides.
  18. [18]
    [PDF] A Path Independent Integral and the Approximate Analysis of Strain ...
    A line integral is exhibited which has the same value for all paths surrounding the tip of a notch in the two-dimensional strain field of an elastic or ...
  19. [19]
    [PDF] FRACTURE MECHANICS - faculty.fairfield.edu
    This threshold value of K is called the fracture toughness or critical stress intensity factor, KIC, and is a characteristic of the material measured by testing ...
  20. [20]
    Overview > Linear Elastic Fracture Mechanics (LEFM)
    ... write the crack propagation criterion in mode I as: KI{<KIC→ the crack does not propagate,≥KIC→ the crack does propagate,. where KIC is the mode I toughness.Missing: K_c | Show results with:K_c
  21. [21]
    E399 Standard Test Method for Linear-Elastic Plane-Strain Fracture ...
    Aug 16, 2019 · 5.1 The property KIc determined by this test method characterizes the resistance of a material to fracture in a neutral environment in the ...
  22. [22]
    Fracture toughness of a metal–organic framework glass - PMC
    May 22, 2020 · KIc is determined to be ~0.1 MPa m0.5, which is even lower than that of brittle oxide glasses due to the preferential breakage of the weak ...
  23. [23]
    Solved Typical values of fracture toughness for various | Chegg.com
    Dec 18, 2020 · Mild steel has 140 MPa, medium-carbon steel 51 MPa, rotor steels 204-214 MPa, and high-strength steels 50-154 MPa fracture toughness.
  24. [24]
    THE EFFECT OF STRAIN RATE ON FRACTURE TOUGHNESS OF ...
    The initiation toughness decreases as the strain rate increases. Note that the stress intensity factors as well as the R-curve slope for 0.08 s−1 are much ...
  25. [25]
    Fracture Toughness - Nondestructive Evaluation Physics : Materials
    At this point the value of KI becomes relatively constant and this value, KIC, is a true material property which is called the plane-strain fracture toughness.
  26. [26]
    Fracture Criteria - Engineering Library
    Consequently the fracture toughness (Kc) may be obtained by the use of crack growth resistance curves (commonly called R-curves).
  27. [27]
    Griffith Theory - an overview | ScienceDirect Topics
    The Griffith theory is defined as the concept stating that, in solids, the formation or extension of a crack requires an increase in potential energy ...
  28. [28]
    [PDF] The Phenomena of Rupture and Flow in Solids - AA Griffith
    The average value of R √e is 239, and the maximum 251. According to the theory, fracture should not depend on Q, and R √c should have, at fracture, the constant ...
  29. [29]
    [PDF] Review of Fracture Toughness (G, K, J, CTOD, CTOA) Testing and ...
    The concept of fracture resistance was first introduced by Irwin and Kies [39] using the energy approach. ... The original energy release rate is defined ...
  30. [30]
    [PDF] YIELDING OF STEEL SHEETS CONTAINING SLITS
    This point has been discussed by FROST and DUGDALE (1958). If the sheet is to yield when the tensile stress at any point reaches the uniaxial yield stress, it ...
  31. [31]
    [PDF] Bearing Pressures and Cracks - Polymerservice Merseburg
    By H. M. WESTERGAARD,¹ CAMBRIDGE, MASS. The task is undertaken of determining the bearing pressures, and the stresses and deformations created by them, in some ...
  32. [32]
    Analysis of Stresses and Strains Near the End of a Crack Traversing ...
    Discussion of this paper should be addressed to the Secretary,. ASME, 29 West 39th Street, New York, N. Y„ and will be accepted until October 10, 1957, for ...
  33. [33]
    [PDF] Stress-Analysis-of-Cracks-Tada-Paris-Irwin.pdf - rezaabedi.com
    Feb 3, 2025 · NOTE: 1. Load is applied along the centerline of the strip at the crack location for uniform pressure on crack surfaces).
  34. [34]
    3 Asymptotic approximations to crack problems
    Asymptotic methods play an important role in all branches of applied mathematics in the evaluation of the solutions of problems depending on a.
  35. [35]
    The distribution of stress in the neighbourhood of a crack ... - Journals
    The effect of a crack (of either type) on the stress produced in an elastic body by a uniform tensile stress is considered and the conditions for rupture ...
  36. [36]
    E399 Standard Test Method for Linear-Elastic Plane-Strain Fracture ...
    Jul 11, 2023 · This test method characterizes the resistance of a material to fracture in a neutral environment in the presence of a sharp crack under essentially linear- ...
  37. [37]
    Improved Elastic Compliance Equation and its Inverse Solution for ...
    Oct 26, 2017 · Compact tension ... Compact tension (CT) specimen is one of the most often used standard specimens with crack length ratios of 0.45≤a/W≤0.70 ...
  38. [38]
    ASTM E399: stress intensity factor K1C - ZwickRoell
    K1C determination is normally performed on brittle materials. A defined crack is first created in the specimen by pre-cracking it in accordance with ASTM E399.
  39. [39]
    [PDF] stress-intensity factors for three-point bend specimens by boundary ...
    The values of Y as a function of relative crack depth a/W for three-point bending are appre- ciably lower than the corresponding values for pure bending ( ...
  40. [40]
    Wide range stress intensity factor expressions for ASTM E 399 ...
    Wide range stress intensity factor expressions for ASTM E 399 standard fracture toughness specimens. Int J Fract 12, 475–476 (1976).
  41. [41]
    On the measurement of ceramic fracture toughness using single ...
    The single edge notched beam (SENB) method is a popular method for measuring the fracture toughness of ceramics. However, standard 3-point or 4-point bend ...
  42. [42]
    Crack front straightness qualification in sent specimens - TWI
    There are no standardised crack front straightness requirements for SENT specimens, so equivalent requirements for SENB specimens are applied, but not often met ...Missing: limitations | Show results with:limitations
  43. [43]
    ASTM D5528 | ISO 15024 GIc Test / DCB Test - ZwickRoell
    The G1c test according to ASTM D5528 or ISO 15024 can be used to determine the critical interlaminar energy release rate of composite laminates.
  44. [44]
    The stress intensity factor for the double cantilever beam
    With terms of order h-squared/a-squared retained in the series expansion, the dimensionless stress intensity factor is found to be K (h to the 1/2)/P = 12 to ...Missing: formula ASTM standard
  45. [45]
    (PDF) On the Analysis and Design of the End Notched Flexure (ENF ...
    The end notched flexure (ENF) specimen is examined as a candidate for measuring interlaminar fracture toughness in skew symmetric loading.
  46. [46]
    ISO 15114:2014 - Fibre-reinforced plastic composites
    In stockISO 15114:2014 specifies a method for the determination of mode II shear load delamination resistance. GIIC, (critical energy release rate), of unidirectional ...
  47. [47]
    Stress intensity factors for a centrally cracked Brazilian disk under ...
    Based on the formulae for the stress field, the analytic solutions of the mode I and mode II stress intensity factors for a Brazilian disk with central crack ...
  48. [48]
    Formula for calculating the critical stress intensity factor in rock ...
    Aug 10, 2025 · Formula for calculating the critical stress intensity factor in rock fracture toughness tests using cracked chevron notched Brazilian disc ( ...
  49. [49]
    D6671/D6671M Standard Test Method for Mixed Mode I-Mode II ...
    Mar 4, 2022 · 1.1 This test method covers the determination of interlaminar fracture toughness, Gc, of continuous fiber-reinforced composite materials at ...
  50. [50]
    Update: Application of the Finite Element Method to Linear Elastic ...
    The purpose of this paper is to present an accurate and efficient method for calculating stress intensity factors for mixed mode deformation.Missing: seminal | Show results with:seminal
  51. [51]
    Triangular quarter‐point elements as elastic and perfectly‐plastic ...
    The proposed elements have wide application in the fracture analysis of structures where ductile fracture is investigated. They permit a determination of the ...
  52. [52]
    Stress‐intensity factor computation in three dimensions with quarter ...
    A consistent method for computing stress-intensity factors from three-dimensional quarter-point element nodal displacements is presented.
  53. [53]
    [PDF] the quarter-point quadratic isoparametric element as a singular ...
    The stress intensity factors. K I and KII are computed at the quarter points using the Westergaard near field displacements which, for plane stress, are given.
  54. [54]
    Analytical and finite element comparisons of stress intensity factors ...
    In this paper, a comparative study regarding the determination of stress intensity factors (SIF) for non-standard thin composite plates is presented.Missing: seminal | Show results with:seminal
  55. [55]
    Boundary Element Analysis in Computational Fracture Mechanics
    Free delivery 14-day returnsThe focus of the monograph is on the application of EIE modeling to one of the most important of the solid mechanics disciplines--fracture mechanics.
  56. [56]
    [PDF] Boundary element formulations in fracture mechanics - WIT Press
    The Boundary Element Method (BEM) is used for stress intensity factors and crack growth analysis in linear elastic fracture mechanics (LEFM). It is also used ...
  57. [57]
    [PDF] A review of SIF evaluation and modelling of singularities in BEM
    To satisfy the continuity requirements of the hypersingular integrals in the traction and flux equa- tions, all the elements on the crack edges are modelled as.
  58. [58]
    Boundary Element Formulations in Fracture Mechanics
    This article reviews advances in the application of boundary element methods (BEM) to fracture mechanics which have taken place over the last 25 years.<|control11|><|separator|>
  59. [59]
    Why the Boundary Element Method is ideal for fracture mechanics?
    Apr 23, 2018 · The BEM solution is well suited for dynamic fracture problem as the infinite domain is easily incorporated in integral solution. 5. The solution ...
  60. [60]
    Fracture & Crack Growth - BEASY Software
    This software is used by engineers, performing damage-tolerant design assessments, to determine accurate stress intensity factor (SIF) solutions and simulate 3D ...Missing: FRANC3D calculation
  61. [61]
    An advanced fast multipole dual boundary element method for ...
    Nov 25, 2024 · A new fast multipole dual boundary element method (FMDBEM) is developed to analyze multiple crack propagations.
  62. [62]
    Adaptive scaled boundary finite element method for fracture in ...
    This study develops a numerical framework for thermo-poroelastic fracture analysis using the Scaled Boundary Finite Element Method (SBFEM) and adaptive ...Missing: advances | Show results with:advances
  63. [63]
    (PDF) Symmetric FEM-driven BEM-FEM coupling for 3D linear ...
    The BEM has strengths in handling problems with stress concentrations. Coupling both approaches is a nat- ural way of combining the strengths of each. In ...<|control11|><|separator|>
  64. [64]
    [PDF] AC 25.571-1D - Advisory Circular - Federal Aviation Administration
    Jan 13, 2011 · damage-tolerance and fatigue evaluation of transport category aircraft structure, including evaluation of widespread fatigue damage (WFD) and ...Missing: intensity | Show results with:intensity
  65. [65]
    [PDF] Development of safety factors to be used for evaluation of cracked ...
    The values of SF\ originate from ASME XI, IWB-3612 where the safety factor in terms of the stress intensity factor and fracture toughness is VWor A/2 depending ...
  66. [66]
    [PDF] AAR-89-03 - NTSB
    Jun 14, 1989 · On April 28, 1988, at 1346, a Boeing 737-200, N73711, operated by. Aloha Airlines Inc., as flight 243, experienced an explosive decompression.
  67. [67]
    [PDF] nasgro 3.0 - a software for analyzing aging aircraft
    There are three basic crack growth models in NASGRO for describing the growth behavior of cracks subjected to cyclic loads. They are: 1) NASGRO equation 2) ...
  68. [68]
    None
    ### Key Assumptions and Limitations of Linear Elastic Fracture Mechanics (LEFM) from the Document
  69. [69]
    [PDF] Fracture mechanics – fundamentals
    a, t, (W-a), h & 2.5(K. IC/ σy. )2#. ▫ If not, apply corrections. ▫ If true ... )2. ▫. This is the most usual criterion for static LEFM. ▫. Conservative or ...
  70. [70]
    [PDF] Failure Behavior and Scaling of Graphene Nanocomposites
    Mar 3, 2017 · (D − a) ≥ 2.5(KIC/σy)2. (18) where D is the specimen width, a is the crack length, KIC is the fracture toughness, and σy is the yielding stress.
  71. [71]
    [PDF] FUNDAMENTALS OF LEFM & APPLICATIONS TO FATIGUE ...
    – The stress intensity factor, K. – The fracture toughness, K c . – The applicable fatigue crack growth rate expression.Missing: K_c | Show results with:K_c
  72. [72]
    Elastic Plastic Fracture Mechanics - an overview - ScienceDirect.com
    5.6.​​ In ductile materials, the LEFM approach is no longer adequate to describe the behaviour of cracks, and modifications to the fracture mechanics solutions ...
  73. [73]
    On small scale nonlinearity and nested crack tip fields in a neo ...
    Aug 13, 2025 · Linear elastic fracture mechanics (LEFM) proves inadequate for soft materials due to their large deformations and nonlinear mechanical responses ...
  74. [74]
    A review on phase field models for fracture and fatigue - ScienceDirect
    Sep 1, 2023 · The purpose of this paper is to provide a comprehensive review of recent work on phase field models for fatigue damage and failures.
  75. [75]
    Phase field study of the effective fracture energy increase ... - arXiv
    Mar 10, 2025 · Indeed, the LEFM theory relies on the scale separation between the macroscopic material size (that can be of the order of mm to km) and the ...