Access the full text.
Sign up today, get DeepDyve free for 14 days.
Z Yang (2009)
The stress and strain concentrations of an elliptical hole in an elastic plate of finite thickness subjected to tensile stressInt J Fract, 155
GN Savin (1961)
Stress concentration around holes
E Sternberg M Sadowsky (1949)
Stress concentration around a triaxial ellipsoidal cavityASME J Appl Mech, 16
CE Inglis (1913)
Stresses in a plate due to the presence of cracks and sharp cornersTrans Inst Naval Archit, 55
WD Pilkey (1997)
Peterson’s stress concentration factors
D Taylor D Bellett (2005)
The fatigue behaviour of three-dimensional stress concentrationsInt J Fatigue, 27
M Sadowsky E Sternberg (1949)
Three-dimensional solution for the stress concentration around a circular hole in a plate of arbitrary thicknessASME J Appl Mech, 16
C-B Kim Z Yang (2008)
The concentration of stress and strain in finite thickness elastic plate containing a circular holeInt J Solids Struct, 45
W Guo P Yu (2008)
The influence of Poisson’s ratio on thickness-dependent stress concentration at elliptic holes in elastic platesInt J Fatigue, 30
L Freund W Yang (1985)
Transverse shear effects for through-cracks in an elastic plateInt J Solids Struct, 21
W Guo L Dai (2006)
Stress concentration at an elliptic hole in transversely isotropic piezoelectric solidsInt J Solids Struct, 43
J Goodier (1970)
Theory of elasticity
A Buczynski A Moftakhar (1994)
Calculation of elasto-plastic strains and stresses in notches under multiaxial loadingInt J Fract, 70
W Guo C She (2007)
Three-dimensional stress concentrations at elliptic holes in elastic isotropic plates subjected to tensile stressInt J Fatigue, 29
G Nicoletto P Livieri (2003)
Elastoplastic strain concentration factors in finite thickness platesJ Strain Anal, 38
Wang CH Guo W (1998)
Guo W, Wang CH, Rose L. Elastoplastic analysis of notch-tip fields in strain hardening materials. Report DSTO-RR-0137, DSTO Aeronautical and Maritime Research Laboratory, Melbourne, 1998.
RD Mindlin TR Kane (1956)
High frequency extensional vibrations of platesASME J Appl Mech, 23
P Lazzarin F Berto (2004)
Three-dimensional linear elastic distributions of stress and strain energy density ahead of V-shaped notches in plates of arbitrary thicknessInt J Fract, 127
A Newport G Glinka (1987)
Universal features of elastic notch-tip stress fieldsInt J Fatigue, 9
W Guo Z Li (2000)
Three-dimensional elastic stress fields near notches in finite thickness platesInt J Solids Struct, 37
J-J Wang E Folias (1990)
On the three-dimensional stress field around a circular hole in a plate of arbitrary thicknessComput Mech, 6
CH Wang A Kotousov (2002)
Fundamental solutions for the generalised plane strain theoryInt J Eng Sci, 40
PC Paris M Creager (1967)
Elastic field equations for blunt cracks with reference to stress corrosion crackingInt J Fract Mech, 3
M Sadowsky (1947)
Stress concentration around an ellipsoidal cavity in an infinite body under arbitary plane stress perpendicular to the axis of revolution of cavityASME J Appl Mech, 69
LP Pook (2013)
A 50-year retrospective review of three-dimensional effects at cracks and sharp notchesFatigue Fract Eng Mater Struct, 36
W Guo (1995)
Elasto-plastic three-dimensional crack border field–III. Fracture parametersEng Fract Mech, 51
RG Budynas (2002)
Roark’s formulas for stress and strain
PA Carraro M Zappalorto (2015)
An engineering formula for the stress concentration factor of orthotropic composite platesCompos Part B Eng, 68
C Kirsch (1898)
Die theorie der elastizitat und die bedurfnisse der festigkeitslehreZ Ver Dtsch Ing, 42
P Lazzarin F Berto (2012)
Induced out-of-plane mode at the tip of blunt lateral notches and holes under in-plane shear loadingFatigue Fract Eng Mater Struct, 35
W Guo (1993)
Elastoplastic three dimensional crack border field—II. Asymptotic solution for the fieldEng Fract Mech, 46
P Lazzarin A Kotousov (2010)
Effect of the thickness on elastic deformation and quasi-brittle fracture of plate componentsEng Fract Mech, 77
NI Muskhelishvili (1963)
Some basic problems of the mathematical theory of elasticity
P Lazzarin S Filippi (2002)
Developments of some explicit formulas useful to describe elastic stress fields ahead of notches in platesInt J Solids Struct, 39
W Guo Z Li (2001)
Three-dimensional elastic stress fields ahead of blunt V-notches in finite thickness platesInt J Fract, 107
W Guo (1993)
Elastoplastic three dimensional crack border field—I. Singular structure of the fieldEng Fract Mech, 46
C Kirsch (1898)
797Z Ver Dtsch Ing, 42
K Man C Shin (1994)
A practical method to estimate the stress concentration of notchesInt J Fatigue, 16
C Wang A Kotousov (2002)
Three-dimensional stress constraint in an elastic plate with a notchInt J Solids Struct, 39
JTPd Castro RCdO Góes (2014)
3D effects around notch and crack tipsInt J Fatigue, 62
G Glinka M Singh (1996)
Elastic-plastic stress–strain calculation in notched bodies subjected to non-proportional loadingInt J Fract, 76
CE Inglis (1913)
219Trans Inst Naval Archit, 55
R Tovo P Lazzarin (1996)
A unified approach to the evaluation of linear elastic stress fields in the neighborhood of cracks and notchesInt J Fract, 78
J Thompson R Xu (1996)
Approximate expressions for three—dimensional notch tip stress fieldsFatigue Fract Eng Mater Struct, 19
J Zhao C She (2008)
Three-dimensional stress fields near notches and cracksInt J Fract, 151
Acta Mechanica Solida Sinica, Vol. 32, No. 4, August, 2019, 393–430 ISSN 1860-2134 https://doi.org/10.1007/s10338-019-00091-w Formulization of Three-Dimensional Stress and Strain Fields at Elliptical Holes in Finite Thickness Plates 1 1 Wei Guo Wanlin Guo ( State Key Laboratory of Mechanics and Control of Mechanical Structures, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China) Received 17 December 2018; revision received 30 March 2019; Accepted 4 April 2019; published online 7 May 2019 c The Author(s) 2019 ABSTRACT Stress raisers such as holes are inevitable in structures at which stress concentration occurs and the static as well as fatigue strength of the structures can be significantly weakened. Therefore, to accurately evaluate the stress concentration factor and stress fields at holes is of essential importance for structure design and service life prediction. Although stress and strain concentration and fields at holes in finite thickness plates strongly change with and along the thickness, manuals of stress concentration for engineering design are mainly based on two- dimensional theory and no explicit formula is available even for circular holes in finite thickness plates. Here we obtain for the first time a complete set of explicit formulae for stress and strain concentration factors and the out-of-plane constraint factor at circular as well as elliptical holes in finite thickness plates by integrating comprehensive three-dimensional finite element analyses and available theoretical solutions. The three-dimensional stress distributions ahead of holes can also be predicted by the obtained formulae. With their accuracy and the corresponding applicable range being analyzed and outlined in detail, the formulae can serve as an important fundamental solution for three-dimensional engineering structure design and guideline for developing three- dimensional analytical methods. KEY WORDS Three-dimensional stress and strain concentration, Stress field, Elliptical holes, Finite thickness plate, Explicit solution 1. Introduction Machines or structural components inevitably contain stress raisers such as holes that can serve as fatigue crack initiation points. Accurate evaluation of the stress concentration factor (SCF) and stress fields at the stress raisers becomes the prerequisite for structural design and service life predic- tion. Pioneer analytical solutions for SCFs at circular and elliptical holes in infinite thin plates were obtained by Kirsch [1] and Inglis [2]. Complex variable solutions for the same problem were obtained later [3–5]. Rational solutions were reported for holes and notches in infinite thin plates and empiri- cal formulae were proposed for more practical cases [6–12]. However, these theoretical solutions and empirical formulae were mainly obtained within the two-dimensional (2D) frame. In developing analytical three-dimensional (3D) solutions for SCFs at cavities and holes, outstand- ing efforts have been made since 1940s. Sadowsky and Sternberg [13, 14] obtained the stress distribution Corresponding author. E-mail:wlguo@nuaa.edu.cn 394 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) Fig. 1. Illustration of the problem: a geometric configuration of the plate with an elliptical hole subjected to uniaxial tension; b the coordinates; c the 3D finite element mesh sketch map near a general ellipsoidal cavity in an infinite body. Sternberg et al. [15] reported an approximate solu- tion for the 3D stress distributions ahead of a circular hole in an infinite plate with arbitrary thickness, which was revisited with reference to uniaxial tensile loading [16]. A generalized plate theory consider- ing the thickness effect was proposed by Kane and Mindlin [17], and has been widely applied recently [18–21] with the assumption of uniform through-the-thickness extensional strain in the thickness direc- tion. Although the average thickness effect can be reflected in certain degree, the variations of the SCFs along the thickness direction cannot be predicted in the frame of generalized plate theory. An important achievement was made by Folias and Wang [22]. They presented a general analytical solution for circular holes in finite thickness plates on the basis of Navier’s equation and managed to work out the 3D SCF along the thickness for the first time following an elaborate numerical process. They found that the SCF is sensitive to the plate thickness and Poisson’s ratio and showed that the SCF attains its maximum on the mid-plane and the plane near the free surface for thin plate and thick plate, respectively. With the increasing capability of calculations, numerical analyses have been more and more widely used in investigating the SCFs and stress fields at holes and notches in finite thickness plates. Guo et al. [23–27] investigated the 3D stress fields in notched plates of finite thickness subjected to tensile loads. They found that the SCF along the 3D notch tip depends on notch configuration and the thickness to root radius ratio, and its value on the mid-plane can be 25% higher than that on the free surface and deviate about 8% from the 2D solution. Therefore, the use of 2D results cannot satisfy safe design of modern structures [28]. Later researchers [29–31] presented similar results and revealed that stress and strain concentration factors are not the same along the thickness even in linear elastic Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 395 (a) (b) (c) (d) (e) (f) Fig. 2. a–e The variations of the SCF on the mid-plane (2z/B =0) (K ) versus the dimensionless thickness B/a for t mp t = 1, 0.8, 0.5, 0.3, 0.1; f the distributions of the relative errors of the formula and the FE results 396 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 3. The comparison among the results of Eq. (7), the previous study [40] and the 2D FE analysis (a) (b) (c) (d) Fig. 4. The variations of the SCF K normalized by (K ) along the thickness from the mid-plane (2z/B =0) to the t t mp free surface (2z/B = 1) in different elliptical holes for different B/a: a t =1; b t =0.8; c t =0.5; d t =0.3 Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 397 (a) (b) (c) (d) Fig. 5. The variations of SCF K along the thickness from the mid-plane (2z/B = 0) to the free surface (2z/B =1) in different elliptical holes for different B/a: a t =1; b t =0.8; c t =0.5; d t =0.3 plates. The importance of 3D effect has been highlighted in some review articles [32, 33]. Recently, Berto et al. [34] confirmed the existence of the stress field associated with the out-of-plane mode in U-notches, as well as circular and elliptical holes under in-plane shear loading. They also reported that SCFs are distinctly different between 2D and 3D models for elliptical holes under in-plane shear loading. Zappalorto and Carraro [35] published a useful engineering formula for the theoretical SCF of orthotropic notched plates under tension. This engineering formula needs, as inputs, the material elastic constants and the SCF of the corresponding isotropic case. These experimental and numerical evidences show that 3D stress and strain concentration and fields at holes and notches are of great significance for structural design. However, no explicit formula is available even for the distribution of SCF at a circular hole along the plate thickness in a finite thickness elastic isotropic plate yet. Explicit and practical expressions for 3D stress and strain fields are still of great significance in developing three-dimensional analytical methods and rapid evaluation for engineering practice. In this article, we report for the first time a complete set of explicit formulae for 3D stress and strain concentration and fields at circular and elliptical holes in finite thickness plates subjected to uniaxial tensile load. The accuracy and applicable range of the formulae are outlined in detail to guide applications. 2. Computational Procedure and Modeling 2.1. Definition A large elastic isotropic plate with thickness B containing an elliptical hole, which is subjected to a far-end uniaxial tension, is considered here, as illustrated in Fig. 1a. The elliptical major axis is 2a, minor axis is 2b, and the elliptical shape factor is t = b/a for the hole. The corresponding curvature 2 2 radius is ρ = b /a = at . The width of the plate is 2W . The height of the plate is 2H. 398 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) (d) Fig. 6. The relative errors of the fitted values of K and the FE results of K versus 2z/B for different B/a: a t =1; b t t t =0.8; c t =0.5; d t =0.3 The geometrical parameters of these specimens are as follows: a =7mm, H/a = W/a = 20, B/a = 0, 0.5, 1, 2, 3, 4, 5, 8, 10, 30 and t = 1, 0.8, 0.5, 0.3, 0.1. The origin of the Cartesian coordinate system (x, y, z) is at the center of the elliptical hole and the plate free surfaces are at z = ± B/2. For convenience, the corresponding cylindrical coordinate system (r, θ, z) is also adopted, as shown in Fig. 1a. The SCF K (r = a, θ =0) is K = σ /σ (1) t yy net where σ = σ W/(W − a) is the mean stress of the net section on the ligament and σ is the applied net y y stress at the ends of the plate along the y direction. K on the mid-plane of the plate (2z/B =0) is denoted as (K ) . Since the geometrical parameters 2H and 2W are sufficiently large compared t mp with the elliptical major axis 2a, the SCF for plane stress state defined as (K ) = σ /σ can be t p−σ yy y calculated by the 2D finite element (FE) method. Three parameters are introduced to describe the 3D characteristics of the stress fields in the finite thickness plates. The first parameter is the in-plane stress ratio defined as xx T = (2) yy The second parameter, the out-of-plane stress constraint factor, has been defined by Guo [36–38]as zz T = (3) σ + σ xx yy Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 399 (a) (b) (c) (d) Fig. 7. 3D distributions of the fitted values of K and the FE results of K versus B/a and 2z/B in different elliptical t t holes (0.5 ≤ B/a ≤ 10): a t =1; b t =0.8; c t =0.5; d t =0.3 In an isotropic linear elastic cracked body, T = 0 for the plane stress state and T = ν for the plane z z strain state. For the finite thickness plate, 0 ≤ T ≤ ν, where ν is Poisson’s ratio of the material. The third parameter is the equivalent thickness B of the point on the crack front line empirically defined eq by She et al. [39]: B = B[1 − (2z/B) ] (4) eq 2.2. Finite Element (FE) Modeling Due to symmetry in geometries, only one-eighth of each plate is modeled with finite elements (region x ≥ 0, y ≥ 0, 0 ≤ z ≤ B/2). All analyses are carried out using the software ABAQUS. As shown in Fig. 1c, the 3D mesh is constructed with the eight-node solid linear elements. In order to characterize the through-thickness stress gradient, forty element layers are divided along the half-thickness of the plate, and the thickness of the element layer is gradually increased from the free surface (2z/B =1) to the mid-plane (2z/B = 0), with the bias ratio being 100 (t = 1, 0.8) and 1000 (t =0.5, 0.3, 0.1) in mesh sizing control. Within each layer, the size of forty elements enclosing the surface of the hole ◦ ◦ gradually increases along the circumferential direction from θ =90 to θ =0 , where the bias ratio is 1000 in mesh sizing control. In the x-direction, the first forty elements starting from x = a are 400 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) Fig. 8. The comparison between the fitted results and the 2D FE results of a σ | /σ , b σ | /σ , c T | versus xx θ=0 y yy θ=0 y x θ=0 (r − a)/(at ) gradually enlarged with the bias ratio 1000 to reflect the in-plane stress gradient. In particular, the 2D plane model is adopted for B/a = 0. The 2D FE mesh is constructed with the four-node solid linear 2D elements, with the mesh sizing control strategy similar to the 3D model. These FE models represent a compromise between the required level of mesh refinement and the extensive computation time. The applied stress at the end of the plate σ is 50 MPa. Poisson’s ratio ν =0.33 and Young’s modulus E = 200 GPa. 3. Results and Formularization 3.1. SCF at Elliptical Holes 3.1.1. SCF on the Mid-Plane The SCF K calculated by Eq. (1) is an important parameter scaling the extent of stress raisers around holes. The variations of the SCF on the mid-plane (2z/B =0) (K ) versus the dimensionless t mp plate thickness B/a are plotted in Fig. 2a–e. (K ) is not a monotonic function of dimensionless t mp plate thickness (B/a) for circular hole (t =1,a = ρ), which is similar to the conclusion by Li et al. [23]. However, its peak value (K ) is at about B/a = 2, as shown in Fig. 2a, rather than at t mp-max B/ρ =3 ∼ 8. For t = 1 and 0.8, both (K ) are at around B/a =2. For t =0.5 and 0.3, both t mp-max (K ) are at about B/a = 1. In addition, (K ) is at about B/a =0.5for t =0.1. With t mp-max t mp-max smaller t, the increasing rate becomes higher before reaching the peak value and the decreasing rate becomes higher after reaching the peak value. After decreasing from the peak value, (K ) is stable t mp at about B/a = 10. The difference between (K ) and the value at B/a = 0 increases from 0.18 t mp-max Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 401 (a) (b) (c) (d) Fig. 9. The distributions of in-plane stress ratio T on different plane layers for dimensionless thickness B/a =1: a t =1; b t =0.8; c t =0.5; d t =0.3 to 1.5 with the decrease of elliptical shape factor from t =1 to t =0.1. The SCF on the mid-plane at holes can be about 8% higher than the 2D solution. The relationships among (K ) , B/a and t can be fitted by the FE results. For 0 ≤ B/a ≤ 30 and t mp 0.1 ≤ t ≤ 1, (K ) canbeexpressedas t mp W − a K (B/a, t) = g(B/a, t) (5) t mp where q (t) · (B/a) + q (t) · (B/a)+ q (t) · q (t) 1 2 1 4 g(B/a, t)= (6) (B/a) + q (t) · (B/a)+ q (t) 3 4 3 2 −14.929t +39.136t − 36.519t +15.311, (0.3 ≤ t ≤ 1) q (t)= (7) 1+1.969/t, (0.1 ≤ t< 0.3) 3 2 q (t)=10.816t − 6.3259t − 15.3192t +7.3886, 3 2 q (t)=7.1503t − 12.648t +4.3624t − 0.1671, 3 2 q (t)= −2.5714t +9.5321t − 2.1436t +0.5833. (K ) described with Eq. (5) is shown in Fig. 2a–e. As shown in Fig. 2f, the relative errors of Eq. t mp (5) and the FE results are within ± 0.75% in the range of B/a =1 ∼ 30 and t =0.1 ∼ 1. When B/a approaches zero, q (t) becomes the 2D SCF for infinite plates. As can be seen from Fig. 3,Eq. (7)is in good agreement with the 2D FE results as well as the formula proposed by Dai et al. [40], which is in the form of (K ) =1 + 2/t in the range of 0.01 ≤ t ≤ 1. t p−σ 402 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) (d) Fig. 10. The distributions of in-plane stress ratio T on different plane layers for dimensionless thickness B/a =5: a t =1; b t =0.8; c t =0.5; d t =0.3 3.1.2. SCF Along the Thickness Figure 4 shows the variations of SCF K normalized by (K ) along the thickness from the mid- t t mp plane (2z/B = 0) to the free surface (2z/B = 1) in different elliptical holes for different B/a. The data points are the 3D FE results. For t = 1, 0.8, 0.5, 0.3, the minimum K /(K ) along the thickness is t t mp located in the free surface (2z/B =1) for different B/a. However, the maximum K /(K ) along the t t mp thickness is not always located in the mid-plane (2z/B = 0) or the plane layer near the free surface. K /(K ) attains the maximum on the mid-plane for B/a no more than 3 at holes with t =1 and t t mp 0.8. The maximum K /(K ) is on the mid-plane for B/a no morethan2and1for t =0.5 and 0.3, t t mp respectively. The value of B/a = 3, 2 or 1 is called the transition thickness (B /a)proposedbyYang et al. [29]. They pointed out that the location of the maximum K /(K ) is not in the mid-plane or t t mp the plane layer near the free surface, but shifts gradually between the two, when the plate thickness exceeds the transition thickness. Figure 5 shows the variations of SCF K along the thickness from the mid-plane to the free surface in different elliptical holes for different B/a. The complex distribution of K reveals the inherent complexity of 3D problems. The data points are the 3D FE results, and the dash dot lines are the 2D FE results. The trend of K changes slightly when K is greater than the 2D solution. The K near t t t the free surface drops rapidly when it is smaller than the 2D solution. As shown in Fig. 5d, the K near the mid-plane is not always larger than the 2D solution for t =0.3. Therefore, the quantitative description of K can reduce the risk of just using the SCF measured at the free surface or from the 2D results for engineering design. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 403 (a) (b) (c) (d) Fig. 11. The distributions of in-plane stress ratio T on different plane layers for dimensionless thickness B/a = 10: a t =1; b t =0.8; c t =0.5; d t =0.3 For different elliptical factors t, the relationship among the K /(K ) ,2z/B and B/a can be fitted t t mp by the FE results. For 1 ≤ B/a ≤ 10 and 0.3 ≤ t ≤ 1, the relationship can be expressed in the form of K (2z/B, B/a, t) = f (2z/B, B/a, t) (8) K (B/a, t) t mp where p · (2z/B) + p · (2z/B)+ p 1 2 3 f (2z/B, B/a, t)= (9) 3 2 (2z/B) + p · (2z/B) + p · (2z/B)+ p 4 2 3 3 2 p (t) · (B/a) + p (t) · (B/a) + p (t) · (B/a)+ p (t) 11 12 13 14 p (B/a, t)= (B/a) + p (t) · (B/a)+ p (t) 15 16 3 2 p (t) · (B/a) + p (t) · (B/a) + p (t) · (B/a)+ p (t) 21 22 23 24 p (B/a, t)= (B/a) + p (t) · (B/a)+ p (t) 25 26 3 2 p (t) · (B/a) + p (t) · (B/a) + p (t) · (B/a)+ p (t) 31 32 33 34 p (B/a, t)= (B/a) + p (t) · (B/a)+ p (t) 35 36 3 2 p (t) · (B/a) + p (t) · (B/a) + p (t) · (B/a)+ p (t) 41 42 43 44 p (B/a, t)= (10) (B/a) + p (t) · (B/a)+ p (t) 45 46 The fitted value of K /(K ) described with Eq. (8) and the FE result of K /(K ) are plotted t t mp t t mp in Fig. 4. Equation (8) times Eq. (5)isEq. (11)inthe form of W − a K (2z/B, B/a, t)= f (2z/B, B/a, t) · g(B/a, t) (11) W 404 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) (d) Fig. 12. The distributions of σ | /σ on different plane layers for dimensionless thickness B/a =1: a t =1; b t =0.8; yy θ=0 y c t =0.5; d t =0.3 The solid lines in Fig. 5 represent variations of K described with Eq. (11). The error analysis and the scope of validity of the formulae are discussed in Sect. 3.1.3. The parameters in Eq. (10) are tabled in “Appendix”. From Fig. 5 and Eq. (11), it reads that in the interior of the plate the 3D SCF at the circular hole with t = 1 is within 6% above the 2D solution, while the 3D SCF at the circular hole with t = 1 can be 16.5% lower than the 2D solution at the free surface. For the elliptical hole with t =0.3, the deviations of the 3D results from the 2D solutions are within 5.4% in the interior but can be 29.5% at the free surface. 3.1.3. Error Analysis and Scope of Validity of the Formulae Figure 6 shows the relative error of the fitted value of K and the FE result of K versus 2z/B for t t different B/a changing from 1 to 10. The relative error is within ± 1.25% in the range of 0 ≤ 2z/B ≤ 0.95 for elliptical factor t = 1; meanwhile, it is within ± 1.25% in the range of 0 ≤ 2z/B ≤ 0.86 and 0 ≤ 2z/B ≤ 0.96 for t =0.8 and 0.3, respectively. When it comes to t = 1 and 0.3, the fitted value of K has the accuracy of over 5% just near the free surface only for B/a = 10 and 8. In particular, the relative error is within ± 1% in the range of 0 ≤ 2z/B ≤ 1for t =0.5. Figure 7 shows 3D distributions of the fitted value of K versus B/a and 2z/B in different elliptical holes. The surface represents Eq. (11), and the solid lines are plotted according to the FE results. In particular, the FE results at 2z/B =1 for B/a ranging from 0.5 to 10 are marked as black dots at the ends of the solid lines. Accuracy of Eq. (11) decreases near the free surface for t = 1, 0.8, 0.5 when B/a =0.5. Equation (11) is in good agreement with the FE results for t =0.3. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 405 (a) (b) (c) (d) Fig. 13. The distributions of σ | /σ on different plane layers for dimensionless thickness B/a =5: a t =1; b t =0.8; yy θ=0 y c t =0.5; d t =0.3 3.2. In-Plane Stress Ratio T and In-Plane Stress Components σ , σ at Elliptical Holes x yy xx 3.2.1. T , σ , σ in 2D Plates x yy xx Previous studies on constructing approximate formulae for stress distribution ahead of a notch tip have been intensively investigated [8, 41, 42]. The best-known expression is the classical solution [5]to the stress fields in an infinite plate containing an elliptical hole under uniaxial loading condition. This solution is in the form of Eqs. (12)–(14), when SCF is introduced in 2D cases: −2 −4 xx (K ) r − a r − a t p−σ θ=0 = 1+ − 1+ (12) σ 2 ρ ρ −2 −4 yy (K ) 1 r − a 3 r − a t p−σ θ=0 = 1+ 1+ + 1+ (13) σ 3 2 ρ 2 ρ It is easy to derive from Eqs. (12) and (13) that −2 −4 r−a r−a 3 1+ − 1+ ρ ρ xx T | = = (14) θ=0 −2 −4 yy r−a r−a θ=0 2+ 1+ +3 1+ ρ ρ It should be pointed out that Eq. (14) is valid in a large range of (r −a)/ρ for the circular hole, but only within a small distance near the elliptical hole or notch root such as (r − a)/ρ ≤ 0.3[23]. Analogous to the forms of Eqs. (12)–(14), Eqs. (15)–(17) are obtained for elliptical holes with t ranging from 1 to 406 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 14. The distributions of σ | /σ on different plane layers for dimensionless thickness B/a = 10: a t =1; b yy θ=0 y t =0.8; c t =0.5; d t =0.3 0.3 by fitting of the 2D FE analysis results with sufficiently high accuracy. m(t) n(t) (K ) σ | t r − a r − a xx p−σ θ=0 = 1+ − 1+ (15) 2 2 σ 2 at at m(t) n(t) (K ) σ | t M (t) − 1 r − a M (t) − 1 r − a yy p−σ θ=0 = 1+ 1+ +3 1+ (16) 2 2 σ M (t) 4 at 4 at m(t) n(t) r−a r−a M (t) 1+ − 1+ 2 2 σ at at xx T | = = (17) θ=0 m(t) n(t) M (t)−1 M (t)−1 r−a r−a yy θ=0 2+ 1+ +3 1+ 2 2 2 at 2 at 3 2 where m(t)= −1.27t − 0.73, n(t)= −1.58t − 2.42, M (t)= −10.51t +28.399t − 27.624t +12.735. Equations (15)–(17) degenerate to Eqs. (12)–(14) when t =1. Figure 8 shows the comparison between formulae and the 2D FE results of in-plane stress compo- nents and in-plane stress ratio versus (r − a)/(at ). Equations (15) and (16) are in good agreement with the 2D FE results in a large range of 0 ≤ (r − a)/(at ) ≤ 100. As can be seen from parts (a) and (b) of Fig. 8, parts of the valid scope of (r − a)/(at ) are chosen to distinguish the distributions of FE results as well as the fitted results at different elliptical holes (t = 1, 0.8, 0.5, 0.3). The in-plane stress ratio T in Fig. 8c increases from 0 to its peak value and then decreases gradually to 0 with increasing (r − a)/(at ). The smaller is the elliptical factor t, the larger is the peak value of in-plane stress ratio T . There is an overlap between the 2D FE results for different t in the range of 0 ≤ (r − a)/(at ) ≤ 0.3, which can be described by Eq. (14) for different notch configurations. As shown in Fig. 8c, the fitted results described with Eq. (17) are consistent with the FE results. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 407 Fig. 15. The distributions of σ | /σ on different plane layers for dimensionless thickness B/a =1: a t =1; b t =0.8; xx θ=0 y c t =0.5; d t =0.3 3.2.2. T , σ , σ in Finite Thickness Plates x yy xx Figures 9, 10 and 11 show the distributions of in-plane stress ratio T on different plane layers for dimensionless thickness B/a = 1, 5 and 10, respectively. The distributions of in-plane stress ratio T on different plane layers are very similar for elliptical holes. The 2D solution Eq. (17) can describe the in-plane stress ratio T on different plane layers. BasedonEq. (16) in 2D cases, we introduce the 3D SCF to describe the in-plane stress component σ | /σ on different plane layers in finite thickness plates, which can be expressed as yy θ=0 y σ (2z/B, B/a, t)| K (2z/B, B/a, t) W σ | yy θ=0 t yy θ=0 = (18) σ (K ) W − a σ y t p−σ y where K (2z/B, B/a, t) has been derived in Sect. 3.1.2 and (K ) = σ /σ is in Sect. 2.1. t t p−σ yy y To distinguish the distributions of σ | /σ on different plane layers, parts of the valid scope of yy θ=0 y (r − a)/(at ) are chosen in Figs. 12, 13 and 14. Figures 12, 13 and 14 show the distributions of σ | yy θ=0 on different plane layers for dimensionless thickness B/a = 1, 5, 10, respectively. Equation (18) can predict the distributions of σ | very well on different plane layers except for yy θ=0 thefreesurface (2z/B =1) at t = 1, 0.3 for B/a = 10. The prediction with low accuracy attributes to the relatively large error of Eq. (11) in some cases including but not limited to the description of K at the free surface at t = 1, 0.3 for B/a = 10. Equation (18) multiplied by Eq. (17) leads to σ (2z/B, B/a, t)| σ (2z/B, B/a, t)| xx θ=0 yy θ=0 = · T | (19) x θ=0 σ σ y y 408 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 16. The distributions of σ | /σ on different plane layers for dimensionless thickness B/a =5: a t =1; b t =0.8; xx θ=0 y c t =0.5; d t =0.3 The in-plane stress component σ | /σ on different plane layers in finite thickness plates can be xx θ=0 y expressed as Eq. (19). Figures 15, 16 and 17 show the distributions of σ | /σ on different plane xx θ=0 y layers for dimensionless thickness B/a = 1, 5 and 10, respectively. The solid lines in Figs. 15, 16 and 17 represent Eq. (19), while the dash lines represent the 2D formula of Eq. (15). In particular, the distributions of σ | /σ can be well predicted with Eq. (19) even at the free surface. xx θ=0 y 3.3. Out-of-Plane Stress Constraint Factor T and Out-of-Plane Stress Component σ at Elliptical z zz Holes 3.3.1. T and σ at the Root of Holes z zz The local 3D stress field becomes very strong when an elliptical hole degenerates from a circular hole (t = 1)toacrack(t = 0). The out-of-plane stress constraint factor T defined by Eq. (3)has been extensively studied to characterize the out-of-plane stress component near the crack front. The saturated value of T /ν is reported as 2/3 for the circular hole [14]. However, no explicit formula of saturated value of T /ν versus plate thickness for elliptical holes has been reported. By fitting of the 3D FE analysis results with sufficiently high accuracy, the out-of-plane stress constraint factor normalized by Poisson’s ratio at the root of holes, T /ν, changing with and along the finite thickness at elliptical z0 holes can be described as (B /a) + a(t)(B /a) eq eq T /ν = h(2z/B, B/a, t)= A(t) (20) z0 (B /a) + b(t)(B /a)+ c(t) eq eq where B = B[1 − (2z/B) ] (4) eq Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 409 Fig. 17. The distributions of σ | /σ on different plane layers for dimensionless thickness B/a = 10: a t =1; b xx θ=0 y t =0.8; c t =0.5; d t =0.3 A(t)= (21) t +2 3 2 3 2 a(t)= − 2.2124t +4.5951t − 3.752t +2.0118, b(t)= − 2.386t +5.2084t − 4.6204t +2.1116 and 3 2 c(t)= − 1.0587t +6.9335t +0.1583t +0.0333. Figure 18 shows the variations of the out-of-plane stress constraint factor normalized by Poisson’s ratio at the root of holes of the mid-plane, T /ν, versus B/a for elliptical holes. T /ν increases with z0 z0 the dimensionless plate thickness B/a and tends to be stable when B/a exceeds a certain value, which is usually expressed as the sufficiently large thickness. T /ν becomes stable when B/a attains around z0 10 which is almost independent of the elliptical shape factor t. Equation (20) is in good agreement with the FE results on the mid-plane (2z/B = 0). Figure 19 shows T /ν on the mid-plane (2z/B =0) z0 versus the elliptical factor t in a finite thickness plate with B/a = 10. The smaller is the elliptical shape factor t, the larger is the saturated value of T /ν. This is supported by the FE results in the z0 literature [23, 30]. The variations of the out-of-plane stress constraint factor normalized by Poisson’s ratio at the root of holes, T /ν,versus2z/B for different elliptical shape factors t are plotted in Fig. 20. T /ν attains z0 z0 its maximum on the mid-plane and decreases monotonically from the mid-plane (2z/B =0) to the free surface (2z/B = 1). With respect to B /a rather than 2z/B, T /ν can be well predicted with one curve for a given eq z0 elliptical factor t shown in Fig. 21. Figure 22 shows the differences between the formula values and the FE values of T normalized by Poisson’s ratio versus B /a for different elliptical shape factors t. z0 eq Most of the error values are in the range of ± 2.5%, and the maximum error is smaller than 5%. 410 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 18. The variations of the out-of-plane stress constraint factor normalized by Poisson’s ratio at the root of holes of the mid-plane, T /ν, versus B/a for different elliptical shape factors t: a t =1; b t =0.8; c t =0.5; d t =0.3 z0 Fig. 19. The value of the out-of-plane stress constraint factor normalized by Poisson’s ratio at the root of holes of the mid-plane (2z/B =0), T /ν, versus the elliptical factor t in a finite thickness plate with B/a =10 z0 Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 411 Fig. 20. The variations of the out-of-plane stress constraint factor normalized by Poisson’s ratio at the root of holes, T /ν, versus 2z/B for different elliptical shape factors t: a t =1; b t =0.8; c t =0.5; d t =0.3 z0 According to the definition of T and T , the out-of-plane stress component can be expressed as z x σ = T (T +1)σ (22) zz z x yy When r = a and θ =0, Eq. (23) can be obtained as σ = T σ (23) zz z0 yy AccordingtoEqs.(11) and (20), Eq. (23) can be expressed as σ (B /a) + a(t)(B /a) W zz eq eq = νA(t) · K (2z/B, B/a, t) (24) σ (B /a) + b(t)(B /a)+ c(t) W − a y eq eq Figure 23 shows the variations of the out-of-plane stress component σ normalized by the applied zz stress at the ends of the plate σ versus 2z/B for different elliptical shape factors t. The out-of-plane stress component σ attains its maximum on the mid-plane and decreases monotonically from the zz mid-plane (2z/B = 0) to the free surface (2z/B = 1). The FE results can be well predicted with Eq. (24). Figure 24 shows the differences between the fitted values and the FE results of σ normalized by zz the applied stress at the ends of the plate σ versus 2z/B for different elliptical shape factors t.Most of the error values are in the range of ± 5% for B/a ranging from 1 to 10 for t = 1, 0.8 and 0.5. For t =0.3, the error value is between minus 4% and minus 10% with 2z/B ranging from 0 to 0.6, and is within ± 5% with 2z/B ranging from 0.6 to 1. 3.3.2. T and σ in 3D Affected Zone z zz As is well known, a strong 3D stress field is observed within the radius of half plate thickness from the crack tip in a sufficiently thin plate. For U-notch, the 3D affected zone (T > 0) on the mid-plane z 412 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 21. The variations of the out-of-plane stress constraint factor normalized by Poisson’s ratio at the root of holes, T /ν, versus B /a for different elliptical shape factors t: a t =1; b t =0.8; c t =0.5; d t =0.3 z0 eq is limited to a distance of approximately 3B/8 from the notch root and is almost insensitive to B/a changing from 0.27 to 3.64 [23]. A rational solution was also proposed in this literature in the form of −2 −4 (T ) r − a r − a z mp =1 − 4.35 1+0.628 +4.35 1+0.628 (25) (T ) B B z0 mp where (T ) is T on the mid-plane and (T ) is T at the root of holes of the mid-plane. Equation z mp z z0 mp z (25) needs verifying for larger B/a and different elliptical holes. By fitting of the 3D FE results for B/a = 10, when the concept of equivalent thickness B is introduced, we approximate the distribution eq of T within the 3D affected zone on different plane layers for elliptical holes by m(t) n(t) T r − a r − a =1 − α(t) 1+ β(t) + α(t) 1+ β(t) (26) T B B z0 eq eq 2 2 where α(t)= −0.315t +1.8211t +2.4165, and β(t)=1.4745t − 3.4915t +2.8012. m(t)and n(t)have been proposed in Eq. (16). The maximum root-mean-square error is 0.02712 for different elliptical holes (t = 1, 0.8, 0.5, 0.3). Figure 25 shows the variations of the out-of-plane constraint factor T normalized by T , the out-of- z z0 plane constraint factor at the root of holes, versus (r − a)/B on different plane layers for dimensionless thickness B/a = 10. T /T is maximal at the root of holes and decays with the distance from the root z z0 of holes. Equations (25) and (26) are plotted in Fig. 25 when 2z/B = 0. Equation (25) can well predict the 3D FE results only on the mid-plane at the circular hole (t =1) for B/a = 10. Figure 26 shows the variations of T normalized by T versus (r − a)/B on different plane layers for dimensionless z z0 eq thickness B/a = 10. The 3D FE results can be mostly predicted with one curve described as Eq. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 413 (a) (b) (c) (d) Fig. 22. The differences between the formula values and the FE values of T normalized by Poisson’s ratio versus B /a z0 eq for different elliptical shape factors t. a t =1; b t =0.8; c t =0.5; d t =0.3 (26). The 3D affected zone is approximately limited to a distance of 0.5B from the root of holes for eq B/a = 10. Figure 27 shows the variations of T normalized by T versus (r − a)/B on different plane layers z z0 for dimensionless thickness B/a = 1. When 2z/B =0, Eq. (25) can well predict the 3D FE results on the mid-plane at circular holes (t = 1, 0.8, 0.5, 0.3) for B/a = 1.However, Eq.(26) is not adequate for the prediction of T /T for B/a = 1. To check the scope of Eq. (25) in terms of the parameter z z0 2z/B, Figs. 28 and 29 show the variations of T /T versus (r − a)/B on different plane layers for z z0 eq dimensionless thickness B/a = 1 and 5, respectively. Here, we replace B with B in Eq. (25), which eq is plotted as dash lines in Figs. 28 and 29. As shown in Fig. 28, the 3D affected zone is approximately limited to a distance of 0.375B ,0.4B ,0.45B ,0.5B from the root of holes for t = 1, 0.8, 0.5, eq eq eq eq 0.3, respectively. T /T in the 3D affected zone can be mostly predicted with Eq. (25), in which B z z0 is replaced with B on different plane layers except for the one near the free surface. As shown in eq Fig. 29, the 3D affected zone is approximately limited to a distance of 0.5B from the root of holes eq for B/a = 5. The size of 3D affected zone is almost independent of the elliptical shape factor t ranging from 1 to 0.3. When θ = 0, the out-of-plane stress component σ canbeexpressedas zz σ = T (T +1)σ (22) zz z x yy Substituting Eqs. (17), (18), (20), and (26)intoEq. (22) yields, σ T σ (2z/B, B/a, t)| yy zz z θ=0 = T · · (T +1) · (27) z0 x σ T σ y z0 y 414 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) (d) Fig. 23. The variations of the out-of-plane stress component σ normalized by the applied stress at the ends of the plate zz σ versus 2z/B for different elliptical shape factors t: a t =1; b t =0.8; c t =0.5; d t =0.3 Substituting Eqs. (17), (18), (20), and (25)intoEq. (22) yields, σ (T ) σ (2z/B, B/a, t)| yy zz z mp θ=0 = T · · (T +1) · (28) z0 x σ (T ) σ y z0 mp y As can be seen from Figs. 30 and 31, σ can be well predicted by Eq. (27) on different plane layers zz for B/a = 10 in the 3D affected zone. As shown in Figs. 32 and 33,Eq. (28) is consistent with the 3D FE results in the 3D affected zone. 3.3.3. Stress Triaxiality R The stress triaxiality which dominates ductile fracture is defined as R = (29) where σ is hydrostatic stress and σ is the von Mises equivalent stress m e σ = (1 + T )(σ + σ ) (30) m z xx yy 2 2 2 2 2 2 1/2 σ = √ [(σ − σ ) +(σ − σ ) +(σ − σ ) +6(σ + σ + σ )] (31) e xx yy yy zz zz xx xy yz zx Neglecting σ , σ and σ in the 3D affected zone, the stress triaxiality R for θ = 0 canbeexpressed xy yz zx σ as 1/3(1 + T )(T +1) z x R = (32) 2 2 2 1/2 [(1 − T + T )(T +1) − (1 + 2T − 2T )T ] z z x z x z Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 415 (a) (b) (c) (d) Fig. 24. The differences between the fitted values and the FE results of σ normalized by the applied stress at the ends zz of the plate σ versus 2z/B for different elliptical shape factors t: a t =1; b t =0.8; c t =0.5; d t =0.3 Substituting Eqs. (17), (20), and (25)intoEq. (32) yields, 1/3(1 + T )(T +1) z x R | = (33) B/a=1 2 2 2 1/2 [(1 − T + T )(T +1) − (1 + 2T − 2T )T ] z z x z x z Substituting Eqs. (17), (20), and (26)intoEq. (32) yields, 1/3(1 + T )(T +1) z x R | = (34) B/a=10 2 2 2 1/2 [(1 − T + T )(T +1) − (1 + 2T − 2T )T ] z z x z x z Figure 34 shows the variations of R at the root of holes (r = a)versus2z/B for different B/a. R σ σ decreases monotonically from the mid-plane to the free surface along the thickness direction for given plate dimensionless thickness B/a and elliptical factor t. For given plate dimensionless thickness B/a and plane layer 2z/B, the smaller is the elliptical factor t, the larger is R . Equation (32) can predict the variations of R at the root of holes for different B/a in different elliptical holes (t = 1, 0.8, 0.5, 0.3) with the maximum relative error smaller than 4%. Figures 35 and 36 show the distributions of R on different plane layers for dimensionless thickness B/a = 1 and 10, respectively. With the distance to the root of holes increasing, R increases to its peak value and then decreases for different plane layers in different thickness plates containing an elliptical hole. With smaller elliptical shape factor t, the location of maximum R is much closer to the root of holes. As shown in Fig. 35, the distributions of R on different plane layers are very similar for t =1 and 0.8. Equation (33) can predict the 3D FE results with adequate accuracy for engineering design at elliptical holes (t = 1, 0.8, 0.5). The predicted value of R is greater than the FE results for t =0.5 and 0.3. R attains its peak value when the distance to the root of the hole increases to around 0.52a, σ 416 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) (d) Fig. 25. The variations of the out-of-plane constraint factor T normalized by T , the out-of-plane constraint factor at z z0 the root of holes, versus (r − a)/B on different plane layers for dimensionless thickness B/a = 10: a t =1; b t =0.8; c t =0.5; d t =0.3 0.5a,0.15a,0.12a for t = 1, 0.8, 0.5, 0.3, respectively. As can be seen from Fig. 36, R attains its peak value when the distance to the root of holes increases to around a,0.75a,0.5a,0.25a for t = 1, 0.8, 0.5, 0.3, respectively. Equation (34) cannot provide accurate prediction of R in quantitative terms. Even so, Eq. (34) can describe the general trend of increasing to a peak value and then decreasing. The location of maximum R of Eq. (34) is closer to the root of the hole than the FE results. The distance between the predicted location and the actual location is around 0.25a. 3.4. Strain Concentration Factor and Strain Energy Density at Elliptical Holes 3.4.1. Strain Concentration Factor Along the Thickness The strain concentration factor K (r = a, θ =0) is K = ε /ε (35) ε yy net where ε = σ /E is the mean strain of the net section on the ligament. For the 3D stress state of net net elastic body, the strain components can be denoted as ε = [σ − ν(σ + σ )]. According to Eqs. yy yy zz xx (1)–(3), the following expression can be obtained K /K =1 − v(T + T + T · T ) (36) ε t x z x z The in-plane stress ratio T is zero along the root of holes. From Eq. (36), the distribution of K /K x ε t along the root of holes can be expressed as K /K =1 − vT (37) ε t z0 where T is described in Eq. (20). z0 Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 417 Fig. 26. The variations of the out-of-plane constraint factor T normalized by T , the out-of-plane constraint factor at z z0 the root of holes, versus (r − a)/B on different plane layers for dimensionless thickness B/a = 10: a t =1; b t =0.8; c eq t =0.5; d t =0.3 The strain concentration factor is distinctly different from the stress concentration factor along the thickness of the root of holes in finite thickness plates except at the free surface, as shown in Fig. 37. The K /K at the mid-plane is smaller for larger thickness and smaller elliptical shape factor t. Similar ε t distributions of K /K along the root of the holes for different B/a as well as t were reported [30]. ε t However, no explicit formula is available to describe the relationship between the strain concentration factor and the stress concentration factor yet. The distributions of strain concentration factor over stress concentration factor K /K can be described with Eq. (37). The maximum relative error between Eq. ε t (37) and the FE results is smaller than 0.8%. 3.4.2. Strain Energy Density Along the Thickness In the case of linear elastic bodies, the elastic strain energy densities for uniaxial and multiaxial stress states can be expressed as Eqs. (38) and (39), respectively. U = (38) 2E 1+ ν 3(1 − 2ν) 2 2 U = σ + σ (39) e m 3E 2E 2(1 + ν) U/U = + 3(1 − 2ν)R (40) where R are shown in Eqs. (32)–(34). σ 418 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 27. The variations of the out-of-plane constraint factor T normalized by T , the out-of-plane constraint factor at z z0 the root of holes, versus (r − a)/B on different plane layers for dimensionless thickness B/a =1: a t =1; b t =0.8; c t =0.5; d t =0.3 The variations of U/U at the root of holes versus 2z/B for different elliptical shape factors t are showninFig. 38. U/U decreases monotonically from the mid-plane to the free surface along the thickness direction for given plate dimensionless thickness B/a and elliptical factor t. For given plate dimensionless thickness B/a and plane layer 2z/B, the smaller is the elliptical factor t, the larger is U/U . Equation (40) can predict the variations of U/U at the root of holes for different B/a in 0 0 different elliptical holes (t = 1, 0.8, 0.5, 0.3) with the maximum relative error smaller than 2%. 4. Discussion When t =0.1, the variations of K /(K ) and K along the thickness from the mid-plane (2z/B = t t mp t 0) to the free surface (2z/B =1) for different B/a (B/a =0.5, 1, 2, 3, 4, 5, 8, 10) are shown in Fig. 39. Singular points occur in Eq. (8) at around 2z/B =0.16 and 0.22 for B/a = 1 and 0.5, respectively. The variations of K /(K ) and K along the thickness from the mid-plane to the free surface in t t mp t elliptical holes (t = 1, 0.8, 0.5, 0.3, 0.1) for B/a = 30 are shown in Fig. 40. This is not taken into consideration in the fitting of 3D results because the dimensionless thickness B/a = 30 rarely exists in actual engineering structures. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 419 Fig. 28. The variations of the out-of-plane constraint factor T normalized by T , the out-of-plane constraint factor at z z0 the root of holes, versus (r − a)/B on different plane layers for dimensionless thickness B/a =1: a t =1; b t =0.8; c eq t =0.5; d t =0.3 The stress and strain at the root of holes are necessities for predicting the fatigue life with traditional local strain approaches. An engineering method for strain hardening materials has been developed to predict the elastic-plastic stress–strain distribution ahead of the root of holes based on the elastic solutions [43]. Neuber’s rule and the equivalent strain energy density (ESED) method were also adopted in this engineering method. The explicit formulae for stress and strain concentration factors as well as stress fields mentioned above can be used to modify Neuber’s rule and the ESED method. 420 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) (c) (d) Fig. 29. The variations of the out-of-plane constraint factor T normalized by T , the out-of-plane constraint factor at z z0 the root of holes, versus (r − a)/B on different plane layers for dimensionless thickness B/a =5: a t =1; b t =0.8; c eq t =0.5; d t =0.3 5. Conclusions Based on comprehensive three-dimensional finite element analyses and the existing theoretical research, the three-dimensional distributions of stress and strain concentration factors at circular as well as elliptical holes and the stress fields near the holes are fitted into a set of explicit empirical formulae in finite thickness plates. Necessary error analysis and scope of validity are conducted for these explicit formulae. Equations (5) and (7) are also in good agreement with the available results in the literature. Equation (5) shows that SCF on the mid-plane at holes with different elliptical shape factors t =1 ∼ 0.1 can be about 8% higher than the corresponding 2D solution. Equation (7)shows that SCF (K ) on the mid-plane (2z/B = 0) at elliptical holes becomes the 2D SCF for infinite t mp plates when B/a approaches zero. The 3D SCF K changes above the 2D result with deviation within 6% in the interior of the plate, while it drops rapidly near the free surface and can be 29.5% below the 2D result for t =0.3to1and B/a = 1 to 10. The three-dimensional distributions of in-plane as well as out-of-plane stresses ahead of the holes can also be predicted by the presented formulae efficiently. Therefore, the obtained formulae can serve as useful basic explicit solutions for engineering structure design and guideline for developing three-dimensional analytical methods. The following conclusions can be helpful for application of the formulae. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 421 Fig. 30. The variations of the out-of-plane stress component σ normalized by σ versus (r − a)/B on different plane zz y layers for dimensionless thickness B/a = 10: a t =1; b t =0.8; c t =0.5; d t =0.3 5.1. Stress and Strain Concentration Factors and Stress Fields at the Root of Holes (1) In the range of 0 ≤ B/a ≤ 30, Eqs. (5)–(7) are presented explicitly for SCF (K ) on the mid- t mp plane (2z/B = 0) at elliptical holes (0.1 ≤ t ≤ 1). The difference between the peak value of (K ) t mp and K at B/a = 0 increases from 0.18 to 1.5 with the decrease of elliptical shape factor from t =1 to t =0.1. The relative errors of Eq. (5) and FE results are within ± 0.75% in the range of B/a =1 ∼ 30 and t =0.1 ∼ 1. (2) In the range of 1 ≤ B/a ≤ 10, Eqs. (8)–(11) are presented explicitly for SCF K along the thickness from the mid-plane to the free surface at elliptical holes (t = 1, 0.8, 0.5, 0.3). When it comes to t = 1 and 0.3, the equations have accuracy over 5% just near the free surface only for B/a =10 and 8. In particular, the relative error is within ± 1% in the range of 0 ≤ 2z/B ≤ 1for t =0.5. When t =0.1, singular points occur in Eq. (8) at around 2z/B =0.16 and 0.22 for B/a =1 and 0.5, respectively. For B/a =0.5, the description of Eq. (11) is in good agreement with the finite element results for t =0.3 and the accuracy of Eq. (11) decreases near the free surface for t =1, 0.8, 0.5. This is not taken into consideration in the fitting of 3D results because the dimensionless thickness B/a = 30 rarely exists in actual engineering structures. The details are discussed in the discussion section. Equation (37) is explicitly presented to describe the relationship between the stress concentration factor and the strain concentration factor for different elliptical shape factors (t = 1, 0.8, 0.5, 0.3) with the maximum relative error smaller than 0.8%. 422 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 31. The variations of the out-of-plane stress component σ normalized by σ versus (r − a)/(at ) on different plane zz y layers for dimensionless thickness B/a = 10: a t =1; b t =0.8; c t =0.5; d t =0.3 (3) In the range of 0 ≤ B/a ≤ 30, the out-of-plane stress constraint factor normalized by Poisson’s ratio, T /ν, can be described by Eq. (18), changing from the mid-plane to the free surface at the root of elliptical holes (0.1 ≤ t ≤ 1). With respect to B /a rather than 2z/B, T /ν can be well eq z predicted with one curve for a given elliptical factor t. Most of the error values are in the range of ± 2.5% and the maximum error is within 5%. In the range of 1 ≤ B/a ≤ 10, Eq. (22) is obtained to characterize the out-of-plane stress component σ at the root of holes normalized by the applied zz stress at the ends of the plate σ with errors in the range of ± 5% for elliptical holes (t = 1, 0.8, 0.5) and − 10% ∼ +5% for t =0.3. (4) With increasing distance to the root of the hole, the stress triaxiality R increases to its peak value and then decreases for different plane layers in different thickness plates containing an elliptical hole. Equation (32) can predict the variations of the stress triaxiality R at the root of elliptical hole for different B/a in different elliptical holes (t = 1, 0.8, 0.5, 0.3) with the error within 4%. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 423 Fig. 32. The variations of the out-of-plane stress component σ normalized by σ versus (r − a)/B on different plane zz y layers for dimensionless thickness B/a =1: a t =1; b t =0.8; c t =0.5; d t =0.3 5.2. Stress Fields Within the 3D Affected Zone (1) A set of formulae, Eqs. (15)–(19), is obtained for holes with elliptical shape factor t ranging from 1 to 0.3. The 2D solution of Eq. (17) can describe the in-plane stress ratio T on different plane layers in finite thickness plates. The in-plane stress components σ and σ can be predicted very xx yy well by Eqs. (18) and (19) on different plane layers. The 3D affected zone can be described as the multiple of B on different plane layers. eq (2) The distributions of the out-of-plane constraint factor T normalized by T , the out-of-plane z z0 constraint factor at the root of holes, can be described with Eqs. (25) and (26)for B/a =1 and 10, respectively. The out-of-plane stress component σ can be well predicted by Eqs. (28)and zz (27)for B/a = 1 and 10, respectively. 424 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 33. The variations of the out-of-plane stress component σ normalized by σ versus (r − a)/(at ) on different plane zz y layers for dimensionless thickness B/a =1: a t =1; b t =0.8; c t =0.5; d t =0.3 (3) When B/a =1, R attains its peak value when the distance to the root of the hole increases to around 0.52a,0.5a,0.15a,0.12a for t = 1, 0.8, 0.5, 0.3, respectively. Equation (33) can predict the 3D finite element results of R with adequate accuracy for engineering design at elliptical holes (t = 1, 0.8, 0.5). When B/a = 10, R attains its peak value when the distance to the root of the hole increases to around a,0.75a,0.5a,0.25a for t = 1, 0.8, 0.5, 0.3, respectively. Equation (34) can describe the general trend of increasing to the peak value of stress triaxiality R and then decreasing. However, the location of maximum R by Eq. (34) is closer to the root of the hole than the finite element results. The distance between the predicted location and the actual location is around 0.25a. Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 425 Fig. 34. The variations of the stress triaxiality R at the root of holes versus 2z/B for different elliptical shape factors t: a B/a =1; b B/a =5; c B/a =10 Fig. 35. The distributions of the stress triaxiality R on different plane layers for dimensionless thickness B/a =1: a t =1; b t =0.8; c t =0.5; d t =0.3 426 ACTA MECHANICA SOLIDA SINICA 2019 Fig. 36. The distributions of the stress triaxiality R on different plane layers for dimensionless thickness B/a = 10: a t =1; b t =0.8; c t =0.5; d t =0.3 Fig. 37. The distributions of strain concentration factor over the stress concentration factor K /K along the thickness ε t from the mid-plane (2z/B = 0) to the free surface (2z/B = 1) in different elliptical holes for different B/a: a t =1; b t =0.8; c t =0.5; d t =0.3 Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 427 Fig. 38. The variations of U/U at the root of holes versus 2z/B for different elliptical shape factors t: a B/a =1; b B/a =5; c B/a =10 (a) (b) Fig. 39. The variations of a SCF K normalized by (K ) ; b SCF K along the thickness from the mid-plane (2z/B =0) t t mp t to the free surface (2z/B = 1) in the elliptical hole (t =0.1) for different B/a (B/a =0.5, 1, 2, 3, 4, 5, 8, 10) 428 ACTA MECHANICA SOLIDA SINICA 2019 (a) (b) Fig. 40. The variations of a SCF K normalized by (K ) ; b SCF K along the thickness from the mid-plane (2z/B =0) t t mp t to the free surface (2z/B = 1) in elliptical holes (t = 1, 0.8, 0.5, 0.3, 0.1) for B/a =30 Acknowledgements. This work was supported by the National Natural Science Foundation of China (51535005, 51472117), the Research Fund of State Key Laboratory of Mechanics and Control of Mechanical Structures (MCMS- I-0418K01, MCMS-I-0418Y01, MCMS-0417G02, MCMS-0417G03), the Fundamental Research Funds for the Central Universities (NP2017101, NC2018001), and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions. The authors would like to thank Dr. Chongmin She for helpful discussions. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Appendix See Table 1. Table 1. Parameters in Eq. (10) Parameters Values t 1 0.8 0.5 0.3 0.1 p 3.637 4.413 5.567 6.366 7.498 p − 28.99 − 34.43 − 43.75 − 53.98 − 61 p 98.97 109.5 151.2 199.8 195.9 p − 133.8 − 123.9 − 193.2 − 311.2 − 302.4 p − 5.442 − 4.489 − 2.97 − 2.542 − 2.204 p 8.215 5.583 4.844 5.533 2.438 p − 7.632 − 8.605 − 11.28 − 12.79 − 14.96 p 57.03 57.31 74.46 96.09 113.6 p − 205.1 − 174.2 − 271 − 381 − 387.3 p 272.3 180 270 510.6 561.3 p − 5.555 − 4.551 − 2.285 − 2.23 − 2.182 p 9.044 5.82 4.71 7.316 2.962 p 3.019 3.908 5.785 6.972 7.447 p − 10.51 − 18.34 − 29.33 − 32.86 − 51.13 p 3.414 34.82 121.5 165.7 187.8 p 25.83 6.656 − 53.26 − 67.21 − 230.1 p − 5.856 − 4.896 − 1.026 1.55 − 2.063 p 9.322 6.945 2.721 6.575 5.078 p 3.575 4.381 5.556 6.351 7.498 p − 28.88 − 35.19 − 44.7 − 54.98 − 62.02 p 98.46 113.1 154.2 202.5 198.3 p − 131.6 − 127.7 − 196.3 − 315.2 − 304.7 p − 5.461 − 4.573 − 3.012 − 2.601 − 2.21 p 8.207 5.785 5.009 5.676 2.453 46 Vol. 32, No. 4 W. Guo, W. Guo: Formulization of Three-Dimensional Stress and Strain Fields 429 References [1] Kirsch C. Die theorie der elastizitat und die bedurfnisse der festigkeitslehre. Z Ver Dtsch Ing. 1898;42:797– [2] Inglis CE. Stresses in a plate due to the presence of cracks and sharp corners. Trans Inst Naval Archit. 1913;55:219–41. [3] Muskhelishvili NI. Some basic problems of the mathematical theory of elasticity. Groningen: Noordhoff; [4] Savin GN. Stress concentration around holes. New York: Pergamon Press; 1961. [5] Timoshenko S, Goodier J. Theory of elasticity. New York: McGraw Hill; 1970. [6] Shin C, Man K, Wang C. A practical method to estimate the stress concentration of notches. Int J Fatigue. 1994;16(4):242–56. [7] Moftakhar A, Buczynski A, Glinka G. Calculation of elasto-plastic strains and stresses in notches under multiaxial loading. Int J Fract. 1994;70(4):357–73. [8] Lazzarin P, Tovo R. A unified approach to the evaluation of linear elastic stress fields in the neighborhood of cracks and notches. Int J Fract. 1996;78(1):3–19. [9] Singh M, Glinka G, Dubey R. Elastic-plastic stress–strain calculation in notched bodies subjected to non-proportional loading. Int J Fract. 1996;76(1):39–60. [10] Pilkey WD. Peterson’s stress concentration factors. 2nd ed. New York: Wiley; 1997. [11] Young WC, Budynas RG. Roark’s formulas for stress and strain. New York: McGraw-Hill; 2002. [12] Filippi S, Lazzarin P, Tovo R. Developments of some explicit formulas useful to describe elastic stress fields ahead of notches in plates. Int J Solids Struct. 2002;39(17):4543–65. [13] Sadowsky M. Stress concentration around an ellipsoidal cavity in an infinite body under arbitary plane stress perpendicular to the axis of revolution of cavity. ASME J Appl Mech. 1947;69:191–201. [14] Sadowsky M, Sternberg E. Stress concentration around a triaxial ellipsoidal cavity. ASME J Appl Mech. 1949;16(2):149–57. [15] Sternberg E, Sadowsky M, Chicago I. Three-dimensional solution for the stress concentration around a circular hole in a plate of arbitrary thickness. ASME J Appl Mech. 1949;16(1):27–38. [16] Xu R, Thompson J, Topper T. Approximate expressions for three—dimensional notch tip stress fields. Fatigue Fract Eng Mater Struct. 1996;19(7):893–901. [17] Kane TR, Mindlin RD. High frequency extensional vibrations of plates. ASME J Appl Mech. 1956;23:277– [18] Yang W, Freund L. Transverse shear effects for through-cracks in an elastic plate. Int J Solids Struct. 1985;21(9):977–94. [19] Kotousov A, Wang C. Three-dimensional stress constraint in an elastic plate with a notch. Int J Solids Struct. 2002;39(16):4311–26. [20] Kotousov A, Wang CH. Fundamental solutions for the generalised plane strain theory. Int J Eng Sci. 2002;40(15):1775–90. [21] Berto F, Lazzarin P, Wang CH. Three-dimensional linear elastic distributions of stress and strain energy density ahead of V-shaped notches in plates of arbitrary thickness. Int J Fract. 2004;127(3):265–82. [22] Folias E, Wang J-J. On the three-dimensional stress field around a circular hole in a plate of arbitrary thickness. Comput Mech. 1990;6(5–6):379–91. [23] Li Z, Guo W, Kuang Z. Three-dimensional elastic stress fields near notches in finite thickness plates. Int J Solids Struct. 2000;37(51):7617–32. [24] Li Z, Guo W. Three-dimensional elastic stress fields ahead of blunt V-notches in finite thickness plates. Int J Fract. 2001;107(1):53–71. [25] Livieri P, Nicoletto G. Elastoplastic strain concentration factors in finite thickness plates. J Strain Anal. 2003;38(1):31–6. [26] She C, Guo W. Three-dimensional stress concentrations at elliptic holes in elastic isotropic plates subjected to tensile stress. Int J Fatigue. 2007;29(2):330–5. [27] Yu P, Guo W, She C, Zhao J. The influence of Poisson’s ratio on thickness-dependent stress concentration at elliptic holes in elastic plates. Int J Fatigue. 2008;30(1):165–71. [28] Bellett D, Taylor D, Marco S, Mazzeo E, Guillois J, Pircher T. The fatigue behaviour of three-dimensional stress concentrations. Int J Fatigue. 2005;27(3):207–21. [29] Yang Z, Kim C-B, Cho C, Beom HG. The concentration of stress and strain in finite thickness elastic plate containing a circular hole. Int J Solids Struct. 2008;45(3):713–31. [30] Yang Z. The stress and strain concentrations of an elliptical hole in an elastic plate of finite thickness subjected to tensile stress. Int J Fract. 2009;155(1):43–54. 430 ACTA MECHANICA SOLIDA SINICA 2019 [31] G´ oes RCdO, Castro JTPd, Martha LF. 3D effects around notch and crack tips. Int J Fatigue. 2014;62(Sup- plement C):159–70. [32] Kotousov A, Lazzarin P, Berto F, Harding S. Effect of the thickness on elastic deformation and quasi-brittle fracture of plate components. Eng Fract Mech. 2010;77:1665–81. [33] Pook LP. A 50-year retrospective review of three-dimensional effects at cracks and sharp notches. Fatigue Fract Eng Mater Struct. 2013;36:699–723. [34] Berto F, Lazzarin P, Kotousov A, Pook LP. Induced out-of-plane mode at the tip of blunt lateral notches and holes under in-plane shear loading. Fatigue Fract Eng Mater Struct. 2012;35:538–55. [35] Zappalorto M, Carraro PA. An engineering formula for the stress concentration factor of orthotropic composite plates. Compos Part B Eng. 2015;68:51–8. [36] Guo W. Elastoplastic three dimensional crack border field—I. Singular structure of the field. Eng Fract Mech. 1993;46(1):93–104. [37] Guo W. Elastoplastic three dimensional crack border field—II. Asymptotic solution for the field. Eng Fract Mech. 1993;46(1):105–13. [38] Guo W. Elasto-plastic three-dimensional crack border field–III. Fracture parameters. Eng Fract Mech. 1995;51(1):51–71. [39] She C, Zhao J, Guo W. Three-dimensional stress fields near notches and cracks. Int J Fract. 2008;151(2):151–60. [40] Dai L, Guo W, Wang X. Stress concentration at an elliptic hole in transversely isotropic piezoelectric solids. Int J Solids Struct. 2006;43(6):1818–31. [41] Creager M, Paris PC. Elastic field equations for blunt cracks with reference to stress corrosion cracking. Int J Fract Mech. 1967;3(4):247–52. [42] Glinka G, Newport A. Universal features of elastic notch-tip stress fields. Int J Fatigue. 1987;9(3):143–50. [43] Guo W, Wang CH, Rose L. Elastoplastic analysis of notch-tip fields in strain hardening materials. Report DSTO-RR-0137, DSTO Aeronautical and Maritime Research Laboratory, Melbourne, 1998.
"Acta Mechanica Solida Sinica" – Springer Journals
Published: Aug 1, 2019
Keywords: Theoretical and Applied Mechanics; Surfaces and Interfaces, Thin Films; Classical Mechanics
You can share this free article with as many people as you like with the url below! We hope you enjoy this feature!
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.