Lecture 4: Development of Constitutive Equations for Continuum, Beams and Plates

Structural Mechanics 2.080 Lecture 4 Semester Yr Lecture 4: Development of Constitutive Equations for Continuum, Beams and Plates This lecture deals with the determination of relations between stresses and strains, called the constitutive equations. For an elastic material the term elasticity law or the Hooke’s law are often used. In one dimension we would write σ = E (4.1) where E is the Young’s (elasticity) modulus. All types of steels, independent on the yield stress have approximately the same Young modulus E = 2. GPa. The corresponding value for aluminum alloys is E = 0.80 GPa. What actually is σ and  in the above equation? We are saying the “uni-axial” state but such a state does not exist simultaneously for stresses and strains. One dimensional stress state produces three-dimensional strain state and vice versa. 4.1 Elasticity Law in 3-D Continuum The second question is how to extend Eq.(4.1) to the general 3-D state. Both stress and strain are tensors so one should seek the relation between them as a linear transformation in the form σij = Cij,kl kl (4.2) where Cij,kl is the matrix with 9 × 9 = 81 coefficients. Using symmetry properties of the stress and strain tensor and assumption of material isotropy, the number of independent constants are reduced from 81 to just two. These constants, called the Lame’ constants, are denoted by (χ, µ). The general stress strain relation for a linear elastic material is σij = 2µij + λkk δij (4.3) where δij is the identity matrix, or Kronecker “δ”, defined by δij = 1 0 0 0 1 0 0 0 1 or δij = 1 δij = 0 if i = j if i 6= j (4.4) and kk is, according to the summation convention, kk = 11 + 22 + 33 = 4-1 dV V (4.5) Structural Mechanics 2.080 Lecture 4 Semester Yr In the expanded form, Eq. (4.3) reads σ11 = 2µ11 + λ(11 + 22 + 33 ), δ11 = 1 (4.6a) σ22 = 2µ22 + λ(11 + 22 + 33 ), δ22 = 1 (4.6b) σ33 = 2µ33 + λ(11 + 22 + 33 ), δ33 = 1 (4.6c) σ12 = 2µ12 δ12 = 0 (4.6d) σ23 = 2µ23 δ23 = 0 (4.6e) σ31 = 2µ31 δ31 = 0 (4.6f) Our task is to express the Lame’ constants by a pair of engineering constants (E(ν), where ν is the Poisson ratio). For that purpose, we use the virtual experiment of tension of a rectangular bar N 1 2 ε11 Extension positive 3 ε33 Contraction negative N Figure 4.1: Uniaxial tension of a bar. In the conceptual test, measured are the force, displacement and change in the crosssectional dimension. The experimental observations can be summarized as follows: • σ11 is proportional to 11 , σ11 = E11 • 22 is proportional to 11 , 22 = −ν11 • 33 is proportional to 11 , 33 = −ν11 Thus the uniaxial tension is producing the one-dimensional state of stress but three-dimensional state of strain σ11 0 0 11 0 0 σij = ij = 0 22 0 (4.7) 0 0 0 0 0 0 0 0 33 4-2 Structural Mechanics 2.080 Lecture 4 Semester Yr We introduce now the above information into Eq. (4.6). σ11 = 2µ11 + χ(11 − ν11 − ν11 ) = E11 (4.8a) σ22 = 2µ(−ν11 ) + χ(11 − ν11 − ν11 ) = 0 (4.8b) and obtain two linear equations relating (χ, µ) with (E, ν) 2µ + χ(1 − 2ν) = E (4.9a) −2µν + χ(1 − 2ν) = 0 (4.9b) Solving Eq.(4.9) for µ and χ gives E 2(1 − ν) Eν χ= (1 + ν)(1 − 2ν) µ= The general, 3-D elasticity law, expressed in terms of (E, ν) is   E ν σij = ij + kk δij 1+ν 1 − 2ν (4.10a) (4.10b) (4.11) 1 1 The mean stress p where −p = σkk = (σ11 + σ22 + σ33 ) is called the hydrostatic pressure. 3 3 At the same time the sum of the diagonal components of the strain tensor denotes the change of volume. Let us make the so-called “contraction” of the stress tensor in Eq.(4.11), meaning that i = j = k   E ν σkk = kk + kk · 3 (4.12) 1+ν 1 − 2ν where δkk = δ11 + δ22 + δ33 = 1 + 1 + 1 = 3. From the above equations the following relation is obtained between the hydrostatic pressure and volume change −p = κ where κ is the bulk modulus κ= dV V E 3(1 − 2ν) (4.13) (4.14) The elastic material is clearly compressible. It is the crystalline lattice that is compressed but on removal the forces returns to the original volume. The inverted form of the 3-D Hook’s law is ij = 1+ν ν σij − σkk δij E E 4-3 (4.15) Structural Mechanics 2.080 Lecture 4 Semester Yr which in terms of the components yields 11 = 22 = 33 = 12 = 23 = 31 = 1 [σ11 − ν(σ22 + σ33 )] E 1 [σ22 − ν(σ11 + σ33 )] E 1 [σ33 − ν(σ11 + σ22 )] E 1 σ12 2 1 σ23 2 1 σ31 2 (4.16a) (4.16b) (4.16c) (4.16d) (4.16e) (4.16f) E is called the shear modulus. Eq.(4.16) illustrates the coupling of 2(1 + ν) individual direct strains with all direct (diagonal) components of the stress tensor. At the same time there is no coupling in shear response. The shear strain is proportional to the corresponding shear stress. where G = 4.2 Specification to the 2-D Continuum Plane Stress This is the state of stress that develops in thin plates and shells so it requires a careful consideration. The stress state in which σ3j = 0, where the x3 = z axis is in the through thickness direction. The non-zero components of the stress tensor are: σij = σ11 σ12 0 σ21 σ22 0 0 0 0 ij = 11 12 0 21 22 0 0 0 33 (4.17) where i, j = 1, 2, 3 and α, β = 1, 2. Accordingly, σkk = σ11 + σ22 + σ33 = σγγ + 0. The 2-D elasticity law takes the following form in the tensor notation αβ = ν 1+ν σαβ − σγγ δαβ E E (4.18) It can be easily checked from Eq.(4.18) that in plane stress 13 = 23 = 0 but 33 = ν − (σ11 +σ22 ). The through-thickness component of the strain tensor is not zero. Because it E does not enter the plane stress strain-displacement relation, its presence does not contribute to the solutions. It can only be determined afterwards from the known stresses σ11 and σ22 . 1−ν By making contraction kk = σkk , one can easily invert Eq.(4.18) in the form E σαβ = E [(1 − ν)αβ + νγγ δαβ ] 1 − ν2 4-4 (4.19) Structural Mechanics 2.080 Lecture 4 Semester Yr The above equation is a starting point for deriving the elasticity law in generalized quantities for plates and shells. We shall return to that task later in this lecture. Before that, let’s discuss three other important limiting cases E (11 + ν22 ) 1 − ν2 E = (22 + ν11 ) 1 − ν2 E = 12 1+ν σ11 = (4.20a) σ22 (4.20b) σ13 Sheet metal (4.20c) A layer in a thin plate or shell σ22 σ11 σ11 σ22 σ21 σ12 Figure 4.2: Examples of plane stress structures. Plane strain holds whenever 2j = 0. By imposing a constraint on 22 = 0, a reaction immediately develops in the direction as σ22 6= 0. The components of the strain and Eq.(4.8) stress tensors are ij = 11 12 0 21 22 0 0 0 0 σ11 σ12 0 σ21 σ22 0 0 0 σ33 σij = (4.21) Can you show that under the assumption of the plane strain, the reaction stress σ33 = ν(σ11 +σ22 )? The plane strain is encountered in many practical situations, such as cylindrical bending of a plate or wide beam. N 3 3 2 2 1 M 1 N Figure 4.3: Tension of bending of a wide sheet/plate gives rise to plane strain. 4-5 Structural Mechanics 2.080 Lecture 4 Semester Yr Uniaxial Strain Uniaxial strain is achieved when the displacement in two directions are constrained. For example, soil or granular materials are tested in a cylinder (called confinement) with a piston, Fig.(4.4). The uniaxial strain also develops in a compressed layer between two rigid plates. Also high velocity plate-to-plate impact products the one-dimensional strain. Here the only component of the strain tensor is the volumetric strain. The plate-to-plate experiments are conducted to establish the nonlinear compressibility of metals under very high hydrostatic loading σkk = −3p. Similarly, the plane wave in the 3-D space is generating a uniaxial strain. V V Confinement Figure 4.4: Examples of problems in which the strain state is uniaxial. The components of the stress and strain tensor in the uniaxial strain are: σij = σ11 σ12 0 σ21 σ22 0 0 0 σ33 ij = 11 0 0 0 0 0 0 0 0 Where the reaction stresses are related to the active stress σ11 by σ22 = σ33 = (4.22) ν(1 + ν)σ11 . 1 − ν2 Can you prove that? The uniaxial stress state was discussed earlier in this lecture when converting the Lame’ constants into the engineering constants (E, ν). 4.3 Hook’s Law in Generalized Quantities for Beams There are three generalized forces in beams (M, N, ν) but only two generalized kinematic quantities (◦ , κ). There is no generalized displacement on which the shear force could exert work. So the shear force is treated as a reaction in the elementary beam theory. This gives rise to some internal inconsistency in the beam theory, which will be enumerated in a separate section. 4-6 Structural Mechanics 2.080 Lecture 4 Semester Yr The starting point in the derivation of the elasticity law for beams is the Euler-Bernoulli hypothesis, (z) = ◦ + zκ (4.23) and the one-dimensional Hook law, Eq.(4.1), and the definition of the bending moment and axial force in the beam, Eqs.(3.36a-3.36c). Let’s calculate first the axial force N Z Z Z Exx dA = E (◦ + zκ) dA σxx dA = N= AZ AZ AZ Z (4.24) ◦ ◦ z dA dA + Eκ κz dA = E  dA + E =E A A A A ◦ Note that the strain of the middle axis and the curvature of the beam axis are independent of the z-coordinate and could be brought in front of the respective integrals. Also Q = R A z dA is the static (first) moment of inertia of the cross-section. From the definition of the neutral axis, Q = 0. The expression for the axial force reduces then to N = EA◦ (4.25) where EA is called the axial rigidity of the beam. We calculate next the bending moment in a similar way Z Z N= σxx z dA = E(◦ + zκ)z dA A Z A Z (4.26) ◦ 2 = E z dA + Eκ z dA A A Because the first term involving the static moment of inertia vanishes, and the expression for the bending moment becomes M = EIκ (4.27) where EI is called the bending rigidity and Z I= z 2 dA (4.28) A is the second moment of inertia. For the rectangular cross-section (b × h) I= bh3 12 (4.29) The significance of the above derivation is that the bending response is uncoupled from the axial response and vice versa. This property allows to derive the famous stress formula for beams. This is indeed one line derivation   N Mz ◦ σ = E = E( + zκ) = E + EA EI N Mz σ(z) = + (4.30) A I 4-7 Structural Mechanics 2.080 Lecture 4 M z I N/A Semester Yr σ(z) η Figure 4.5: Linear distribution of stresses along the height of the beam. Both axial force and bending moment contribute to the stress distribution along the along the height of the beam, as illustrated in Fig.(4.5). From Eq.(4.30) one can calculate the point z = η where the stresses become zero η=− I N N = −ρ2 AM M (4.31) where ρ is the moment of giration of the cross-section defined by I = ρ2 A. The position of the zero stress axis depends on the ratio of axial force to bending moment. If η < h, where h is the thickness of a rectangular section beam, the zero stress point is inside the beam boundary, there is a bending dominated response. The tension dominated response is when η is several times larger than h. 4.4 Inconsistencies in the Elementary Beam Theory The equations presented in Section 3.6 under the ADVANCED TOPIC were derived without any approximate assumption. In order for the beam to be in equilibrium, shear force V must be present, when the beam is under pure bending (uniform bending over the length of the beam). It is the shear stress σxz that give rise to the shear force, according to the definition, Eq.(3.45). Therefore any inconsistencies must come from the strain-displacement relations as well as constitutive equations, where some approximations were introduced. The presence of the shear stresses σxz = σ13 means that shear strains 13 = xz must develop according to Eq.(4.16). σxz (z) xz (z) = (4.32) 2G The shear strain is defined as   1 ∂ux ∂uz xz = + (4.33) 2 ∂z ∂x The Euler-Bernoulli assumption tells us that the shear strain vanishes. Then, Eq.(4.32) is violated because the LH is zero while the RH is not. Suppose for a while that xz = 0. Then ∂ux ∂w(x) =− = −θ(x) (4.34) ∂z ∂x where uz = w(x) is independent of the coordinate z. Integrating Eq.(4.34) one gets ux (z) = u◦ − zθ 4-8 (4.35) Structural Mechanics 2.080 Lecture 4 Semester Yr which is equivalent to the plane-remain-plane and normal-remain-normal hypothesis, introduced in Lecture 2. Assume now that the out-of-plane strain is a certain given function of z. Performing the integration of Eq.(4.32) in a similar way as before, one gets Z ◦ ux (z) = u − zθ + xz (z) dz (4.36) It transpires from the above results that deformed section are not flat but are warped instead. The amount of warping is given by the third term in Eq.(4.36). Can we estimate the amount of warping? Yes, but we have to go ahead of the presented material and quota the solution for the deflected slope θ of the beam. Le’s settle on the simplest case of a clamped cantilever beam loaded at its tip by the point force P θ= P l2 2EI (4.37) This solution will be derived in Lecture 5. Figure 4.6: Warping of the end section of the cantilever beam. Another result needed is the distribution of shear stresses across the height of the beam. For the rectangular section beam (b × h), the shear stress is a parabolic function of z   z2 3P σxz (z) = 1− (4.38) 2A (h/2)2 The corresponding strain is calculated from Eq.(4.32). Assume that there is no axial force, N = 0, so from Eq.(4.25) ◦ = 0 and u◦ = 0. After integration, the displacement profile defined by Eq.(4.36) becomes   1 3P z3 P l2 ux (z) = − z+ z− (4.39) 2EI 2 2 A 3(h/2)2 In order to quantify the correction of the displacement field due to warping, let’s calculate h the maximum values of the two terms at z = − . The first term arising from the Euler2 Bernoulli assumption gives h ρl2 h uIx (z = − ) = (4.40) 2 2EI 2 4-9 Structural Mechanics 2.080 Lecture 4 Semester Yr The second correction term is 1 Ph h uII x (z = − ) = 2 2G A 2 (4.41) uII E I E  ρ 2 x = = uIx 2G Al2 2G l (4.42) The ratio of the two terms is where ρ is the radius of giration of the cross-section. For a rectangular cross-section (b × h), ρ2 = I bh3 h2 = = A 12bh 12 (4.43) The ratio E/2G is E = 2G E = (1 + ν) E 2 2(1 + ν) Then, the relative amplitude of warping from Eq.(4.42) is   uII 1+ν h 2 x = uIx 12 l (4.44) (4.45) l For a typical beam with = 20, the above ratio becomes 0.25 × 10−3 !!! In order to compare h the plane and wrapped cross-section, the amount of warping had to be magnified thousand times, see Fig.(4.6). It can be concluded that the effect of warping is of an order of 0.1 % and can be safely neglected in the engineering beam theory. In other words the “rein” of the Euler-Bernoulli assumption is unchallenged. Another inconsistency of the elementary beam theory is that the uniaxial stress gives rise to the tri-axial strain state. In particular, from the 3-D constitutive equation, the strain components ν yy = zz = − σxx (4.46) E Let’s take as an example the same cantilever beam with a tip load. The bending moment at root of the beam is M = P l, and from the stress formula, σxx = From the definition yy = Pl z I (4.47) duy , and after integrating with respect to y, one gets dy uy = − The maximum displacement occurs at z = formula (see Lecture 5) δ= P lν zy IE (4.48) h b and y = . Making use of the beam deflection 2 2 P l3 Pl 3δ or = 2 3EI EI l 4-10 (4.49) Structural Mechanics 2.080 Lecture 4 Semester Yr the formula for the maximum displacement of a beam, normalized with respect to the beam thickness becomes    2 (uy )max h 3 δ = ν (4.50) h 4 h l What is the range of the normalized beam deflections δ? The beam deflects elastically until the most stressed fibers reach yield of the materials, σxx z= h = σy . 2 Then, from the stress formula Pl h σy = (4.51) I 2 Combining the above expression with the beam deflection formula, Eq.(4.49), the estimate for the maximum elastic tip displacement   δ 2 σy l 2 = (4.52) h 3E h Combining Eqs.(4.50) and (4.52), the expression for the maximum normalized displacement of the corner of the cross-section becomes (uy )max ν σy = (4.53) h 2E σy 1 With realistic values ν = and = 10−3 , the amount of maximum change of the width of 3 E the beam is 0.1 % of the beam height. Such a tiny change in the cross-sectional dimension has no practical effect on the beam solution. A similar analysis can be performed to estimate the change in the height of the beam. When the signs of z and y coordinates is properly taken into account, the present calculations predict the following change in the shape of the cross-section. (uy)max y x Figure 4.7: Predicted (left) and actual “anticlastic” deformed cross-section of the beam subjected to pure bending. Note that the deflections were magnified by a factor of 104 . The anticlastic deformation can be easily seen by bending a rubber eraser, which is a very short beam. We can conclude the present section that the internal inconsistencies of the beam theory do not produce any significant errors in engineering applications. Therefore, one can safely assume that the cross-section of the beam does not deform and only moves as a rigid body with the increasing beam deflections. 4-11 Structural Mechanics 2.080 Lecture 4 Semester Yr ADVANCED TOPIC 4.5 Derivation of Constitutive Equations for Plates For convenience, the set of equations necessary to derive the elasticity law for plates is summarized below. Hook’s law in plane stress reads: σαβ = E [(1 − ν)αβ + νγγ δαβ ] 1 − ν2 (4.54) In terms of components: E (xx + νyy ) 1 − ν2 E = (yy + νxx ) 1 − ν2 E xy = 1+ν σxx = (4.55a) σyy (4.55b) σxy (4.55c) Here, strain tensor can be obtained from the strain-displacement relations: αβ = ◦αβ + zκαβ (4.56) Now, define the tensor of bending moment: h 2 Z Mαβ ≡ −h 2 σαβ z dz (4.57) σαβ dz (4.58) and the tensor of axial force (membrane force): Z Nαβ ≡ h 2 −h 2 Bending Moments and Bending Energy The bending moment Mαβ is now calculated by substituting Eq.(4.54) with Eq.(4.57) Mαβ = = E 1 − ν2 Z h 2 −h 2 [(1 − ν)αβ + νγγ δαβ ]z dz E [(1 − ν)◦αβ + ν◦γγ δαβ ] 1 − ν2 Z z dz −h 2 E + [(1 − ν)καβ + νκγγ δαβ ] 1 − ν2 = h 2 Z −h 2 Eh3 [(1 − ν)καβ + νκγγ δαβ ] 12(1 − ν 2 ) 4-12 h 2 (4.59) z 2 dz Structural Mechanics Z Note that the term 2.080 Lecture 4 h 2 Semester Yr z dz is zero, as shown in the case of beams. Therefore there are no −h 2 ◦ αβ entering mid-surface strains the moment-curvature relation. Here we define the bending rigidity of a plate D as follows: D= Eh3 12(1 − ν 2 ) (4.60) Now, one gets the moment-curvature relations in the tensorial form Mαβ = D[(1 − ν)καβ + νκγγ δαβ ] M11 M22 M21 M22 Mαβ = (4.61) (4.62) where M12 = M21 due to symmetry. In the expanded notation, M11 = D(κ11 + νκ22 ) (4.63a) M22 = D(κ22 + νκ11 ) (4.63b) M12 = D(1 − ν)κ12 ) (4.63c) One-dimensional Bending Energy Density Here, we use the hat notation for a function of certain argument, such as: M11 = M̂11 (κ11 ) = Dκ11 Then, the bending energy density Ũb reads: Z κ̄ Ūb = M̂11 (κ11 ) dκ11 0 Z κ̄11 =D κ11 dκ11 (4.64) (4.65) 0 1 = D(κ̄11 )2 2 1 Ūb = M11 κ̄11 2 (4.66) General Case General definition of the bending energy density reads: I Ūb = Mαβ dκαβ 4-13 (4.67) Structural Mechanics 2.080 Lecture 4 M11 Semester Yr κ11 κ D κ22 κ11 dκ11 Figure 4.8: In one dimension the energy density is the area under the linear moment-curvature plot. In the multi-axial case the final value can be reached along the straight or nonlinear path. H where the symbol denotes integration along a certain loading path. Let’s calculate the energy density stored when the curvature reaches a given value κ̄αβ along a straight loading path: καβ = ηκ̄αβ (4.68a) dκαβ = κ̄αβ dη (4.68b) Mαβ Mαβ η=1 η η=0 καβ καβ Figure 4.9: The straight loading path in the 3-dimensional space of bending moments.. From the linearity of the moment-curvature relation, Eq.(4.61), it follows that Mαβ = M̂αβ (καβ ) = M̂αβ (ηκ̄αβ ) = η M̂αβ (κ̄αβ ) 4-14 (4.69) Structural Mechanics 2.080 Lecture 4 where M̂αβ (καβ ) is a homogenous function of degree one. I Ūb = M̂αβ (καβ ) dκαβ Z 1 η M̂αβ (κ̄αβ )κ̄αβ dη = 0 Z 1 = M̂αβ (κ̄αβ )κ̄αβ η dη Semester Yr (4.70) 0 1 = M̂αβ (κ̄αβ )κ̄αβ 2 1 = Mαβ κ̄αβ 2 Now, the bending energy density reads D [(1 − ν)κ̄αβ + ν κ̄γγ δαβ ]κ̄αβ 2 D = [(1 − ν)κ̄αβ κ̄αβ + ν κ̄γγ κ̄αβ δαβ ] 2 D = [(1 − ν)κ̄αβ κ̄αβ − ν(κγγ )2 ] 2 Ūb = (4.71) The bending energy density expressed in terms of components are: Ūb = = = = = D {(1 − ν)[(κ̄11 )2 + 2(κ̄12 )2 + (κ̄22 )2 ] + ν(κ̄11 + κ̄22 )2 } 2 D {(1 − ν)[(κ̄11 + κ̄22 )2 − 2κ̄11 κ̄22 + 2(κ̄12 )2 ] + ν(κ̄11 + κ̄22 )2 } 2 D {(κ̄11 + κ̄22 )2 − 2κ̄11 κ̄22 + 2(κ̄12 )2 − ν[−2κ̄11 κ̄22 + 2(κ̄12 )2 ]} 2 D {(κ̄11 + κ̄22 )2 − 2κ̄11 κ̄22 + 2(κ̄12 )2 − ν[−2κ̄11 κ̄22 + 2(κ̄12 )2 ]} 2 D {(κ̄11 + κ̄22 )2 + 2(1 − ν)[−κ̄11 κ̄22 + (κ̄12 )2 ]} 2 Ūb =   D (κ̄11 + κ̄22 )2 − 2(1 − ν) κ̄11 κ̄22 − (κ̄12 )2 2 (4.72) (4.73) The term in the square brackets is the Gaussian curvature, κG , introduced in Lecture 2, Eq.(2.62). Should the Gaussian curvature vanish, as it is often the case in plates, then the bending energy density assumes a very simple form Ūb = 21 D(κ̄11 + κ̄22 )2 . Total Bending Energy The total bending energy is the integral of the bending energy density over the area of plate: Z Ub = Ūb dA (4.74) S 4-15 Structural Mechanics 2.080 Lecture 4 Semester Yr Membrane Forces and Membrane Energy The axial force can be calculated in a similar way as before Z h 2 E [(1 − ν)αβ + νγγ δαβ ] dz Nαβ = 1 − ν2 − h 2 Z h 2 E = [(1 − ν)◦αβ + ν◦γγ δαβ ] dz 1 − ν2 − h 2 Z h 2 E + [(1 − ν)καβ + νκγγ δαβ ]z dz 2 1 − ν −h 2 Z h 2 E ◦ ◦ = [(1 − ν)αβ + νγγ δαβ ] dz 2 1−ν −h 2 Z h 2 E + [(1 − ν)καβ + νκγγ δαβ ] dz 2 1−ν −h (4.75) 2 Eh = [(1 − ν)◦αβ + γ◦γγ δαβ ] 1 − ν2 Z The integral h 2 −h 2 z dz is zero which means that there is no coupling between the membrane force and curvatures. Here we define the axial rigidity of a plate C as follows: C= Eh 1 − ν2 (4.76) Now, one gets the membrane force-extension relation in the tensor notation: Nαβ = C[(1 − ν)◦αβ + ν◦γγ δαβ ] Nαβ = N11 N12 N21 N22 (4.77) (4.78) where N12 = N21 due to symmetry. In components, N11 = C(◦11 + ν◦22 ) (4.79a) C(◦22 (4.79b) N22 = N12 = C(1 + ν◦11 ) − ν)◦11 (4.79c) Membrane Energy Density Using the similar definition used in the calculation of the bending energy density, the extension energy (membrane energy) reads: I Ūm = Nαβ d◦αβ (4.80) 4-16 Structural Mechanics 2.080 Lecture 4 Semester Yr Let’s calculate the energy stored when the extension reaches a given value ¯◦αβ . Consider a straight path: ◦αβ = η¯ ◦αβ d◦αβ = ¯◦αβ dη (4.81a) (4.81b) Nαβ = N̂αβ (◦αβ ) = N̂αβ (η¯ ◦αβ ) (4.82) = η N̂αβ (¯ ◦αβ ) where N̂αβ (◦αβ ) is a homogenous function of degree one. Z ¯◦ αβ Ūm = N̂αβ (◦αβ ) d◦αβ 0 Z 1 ◦αβ dη = η N̂αβ (¯ ◦αβ )¯ 0 (4.83) 1 = N̂αβ (¯ ◦αβ )¯ ◦αβ 2 1 = Nαβ ¯◦αβ 2 Now, the extension energy reads: C ◦γγ δαβ ]¯ ◦αβ [(1 − ν)¯ ◦αβ + ν¯ 2  C ◦γγ )2 = (1 − ν)¯ ◦αβ ¯◦αβ + ν(¯ 2 The extension energy density expressed in terms of components is:  ◦ 2  C Ūm = (1 − ν) (¯ 11 ) + 2(¯ ◦12 )2 + (¯ ◦22 )2 + ν(¯ ◦11 + ¯◦22 )2 2  ◦  C = (1 − ν) (¯ 11 + ¯◦22 )2 − 2¯ ◦11 ¯◦22 + 2(¯ ◦12 )2 + ν(¯ ◦11 + ¯◦22 )2 2   C ◦ = (¯ 11 + ¯◦22 )2 − 2¯ ◦11 ¯◦22 + 2(¯ ◦12 )2 − ν −2¯ ◦11 ¯◦22 + 2(¯ ◦12 )2 2  ◦ ◦  C ◦ = (¯ 11 + ¯◦22 )2 + 2(1 − ν) −¯ 11 ¯22 + (¯ ◦12 )2 2 Ũm = (4.84) (4.85)   C ◦ (¯ 11 + ¯◦22 )2 − 2(1 − ν) ¯◦11 ¯◦22 − (¯ ◦12 )2 (4.86) 2 The total membrane energy is the integral of the membrane energy density over the area of plate: Z Ūm = Um = Ūm dS S END OF ADVANCED TOPIC 4-17 (4.87) Structural Mechanics 4.6 2.080 Lecture 4 Semester Yr Stress Formula for Plates In the section on beams, it was shown that the profile of axial stress can be determined from the known bending moment M and axial force N , see Eq. (4.30). A similar procedure can be developed for plates by comparing Eqs (4.61-4.77) with Eq. (4.54). The stress-strain curve for the plane stress can be expressed in terms of the middle surface strain tensor ◦αβ and curvature tensor καβ by combining Eqs. (4.54) and (4.56). σαβ = E [(1 − ν)◦αβ + ν◦γγ δαβ ] 1 − ν2 E [(1 − ν)καβ + νκγγ δαβ ]z + 1 − ν2 (4.88) From the moment-curvature relation, Eq. (4.61): (1 − ν)καβ + νκγγ δαβ = Mαβ D (4.89) (1 − ν)◦αβ + ν◦γγ δαβ = Nαβ C (4.90) Similarly, from Eq. (4.72) where D = Eh3 Eh is the bending rigidity, and C = is the axial rigidity of the 2 12(1 − ν ) 1 − ν2 plate. From the above system, one gets σαβ = Ez Mαβ E Nαβ + 1 − ν2 D 1 − ν2 C (4.91) zMαβ Nαβ + 3 h h /12 (4.92) or using the definitions of D and C σαβ = The above equation is dimensionally correct, because both Nαβ and Mαβ are respective quantities per unit length. In particular stress in the case of cylindrical bending is σxx = Nxx zMxx + 3 h h /12 (4.93) Multiplying both the numerators and denominators of the two terms above by b yields σxx = Nxx b zMxx b + 3 hb bh /12 (4.94) Now, observing that Nxx b = N is the beam axial force, bMxx = M is the beam bending bh3 moment, hb = A is the cross-section of the rectangular section beam, and is the moment 12 of inertia, the familiar beam stress formula is obtained σ= N Mz + A I 4-18 (4.95)