Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.

ca

2D Elasticity Theory
The 2D elasticity problem is here formulated in the x-y coordinate system using the
notation shown in Figure 1. u(x,y) and v(x,y) are the displacements in the x and y
directions, respectively. Similarly, px(x,y) and py(x,y) are the distributed loads in those
directions.
⎧
⎪ px ⎫
⎪ ⎧ u ⎫
⎨ ⎬ ⎨ ⎬
⎪ py ⎪ ⎩ v ⎭
⎩ ⎭

Kinematic
Equilibrium compatibility

⎧ σ ⎫ ⎧ ε ⎫
⎪⎪ xx ⎪⎪ ⎪⎪ xx ⎪⎪
Material
⎨ σ yy ⎬ law ⎨ ε yy ⎬
⎪ ⎪ ⎪ ⎪
⎪⎩ τ xy ⎪⎭ ⎪⎩ γ xy ⎪⎭
Figure 1: Ingredients of the 2D boundary value problem.

Material Law
Hooke’s law for 2D continuum problems has two versions: plane stress and plane strain.
Plane stress applies when there are no stresses perpendicular to the considered plane. This
applies to thin planar members that do not have stresses imposed perpendicular to the
surface. One example is a deep beam that stretches along the x-axis with z as the vertical
axis. Seen from the side this beam forms a 2D continuum problem in the x-z-plane with
plane stress material law because there is no stress perpendicular to the sides of the beam.
Conversely, plane strain implies that there is zero strain perpendicular to the plane under
consideration. This applies to problems where the plane is a cross-section of a structure
that is long in the direction perpendicular to that plane. One example is the modelling of
the cross-section of a dam structure. From the selected cross-section the dam stretches out
to both sides until it meets mountainside supports. This restrains displacement, and hence
strain, perpendicular to the x-y-plane That plane is then in a state of plane strain, i.e.,
without strain perpendicular to the plane.
Plane Stress
The plane stress version of the material law is the straightforward simplification of the
material law for 3D problems when one coordinate is removed:

2D Elasticity Theory Updated May 22, 2019 Page 1

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

σ xx σ
ε xx = − ν ⋅ yy
E E
σ σ
ε yy = yy − ν ⋅ xx (1)
E E
τ xy = G ⋅ γ xy

( )
where G = E 2 ⋅(1+ ν ) . In Voight notation any 2D material law is written generically
in index and vector notation as follows:
σ i = Dij ⋅ ε j ⇔ σ = Dε (2)

where the D-matrix for plane stress is

⎧ σ ⎫ ⎡ 1 ν 0 ⎤ ⎧ ε xx ⎫
⎪⎪ xx ⎪⎪ ⎢ ⎥ ⎪ ⎪⎪
E ⎢ ν 1 0 ⎥ ⎪ ε
⎨ σ yy ⎬ = ⋅ ⎨ yy ⎬ (3)
⎪ ⎪ 1− ν 2
⎢ 1− ν ⎥ ⎪ ⎪
⎪⎩ σ xy ⎪⎭ ⎢ 0 0 ⎥ ⎪ γ xy ⎪⎭
2
!### ⎣ #"#### $⎦ ⎩
D
or inversely, directly from Eq. (1):
⎧ ε ⎫ ⎤ ⎧⎪ σ xx ⎫
⎪⎪ xx ⎪⎪ 1 ⎡⎢ 1 −ν 0
⎥ ⎪ ⎪⎪
⎨ ε yy ⎬ = ⎢ −ν 1 0 ⎥ ⋅ ⎨ σ yy ⎬ (4)
⎪ ⎪ E⎢ 0 0 2(1+ ν ) ⎥ ⎪ τ ⎪
⎪⎩ γ xy ⎪⎭ ⎣ ⎦ ⎪⎩ xy ⎪⎭

Plane Strain
When there is zero strain perpendicular to the x-y-plane then the material law reads
⎧ σ ⎫ ⎡ 1− ν ν 0 ⎤ ⎧ ε ⎫
⎪⎪ xx ⎪⎪ ⎢ ⎥ ⎪ xx ⎪⎪
⎨ σ yy ⎬=
E ⎢ ν 1− ν 0 ⎥ ⋅ ⎪⎨ ε yy ⎬ (5)
⎪ ⎪ (1+ ν )(1− 2ν ) ⎢ 1− 2ν ⎥ ⎪ ⎪
⎪⎩ σ xy ⎪⎭ ⎢ 0 0 ⎥ ⎪ γ xy ⎪⎭
⎣ 2 ⎦ ⎩
or inversely
⎧ ε ⎫ ⎡ ⎤ ⎧ σ xx ⎫
⎪⎪ 1 ⎢ 1− ν −ν (1+ ν )
2
⎪⎪ xx 0
⎥ ⎪⎪ ⎪⎪
⎨ ε yy ⎬ = ⎢ −ν (1+ ν ) 1− ν 2 0 ⎥ ⋅ ⎨ σ yy ⎬ (6)
⎪ ⎪ E ⎢ ⎥ ⎪ ⎪
⎪⎩ γ xy ⎪⎭ ⎢⎣ 0 0 2(1+ ν ) ⎥⎦ ⎪⎩ σ xy ⎪⎭

2D Elasticity Theory Updated May 22, 2019 Page 2

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

Kinematic Compatibility
The 2D specialization of the 3D kinematic compatibility equations from the general
theory of elasticity is trivial and reads
⎧ ε ⎫ ⎡
⎪⎪ xx ⎪⎪ ⎢ ∂ ∂x 0 ⎤
⎥ ⎪⎧ u ⎪⎫
⎨ ε yy ⎬= ⎢ 0 ∂ ∂y ⎥⎨ ⎬ (7)
⎪ ⎪ ⎢ ∂ ∂ y ∂ ∂x ⎥ ⎩⎪ v ⎭⎪
⎪⎩ γ xy ⎪⎭ ⎣ ⎦

Differentiating the first equation twice with respect to y, then differentiating the second
equation twice with respect to x, and finally differentiating the third equation with respect
to x and y yields
∂ 2 ε xx ∂3 u
= (8)
∂ y2 ∂x ∂ y 2

∂ 2 ε yy ∂3 v
= (9)
∂x 2 ∂ y ∂x 2

∂ 2 γ xx ∂3 u ∂3 v
= 2 + 2 (10)
∂ y ∂x ∂ y ∂x ∂x ∂ y
Substituting Eqs. (8) and (9) into Eq. (10) yields the “compatibility equation” that states
the necessary relationship between the strains for the deformation pattern to be physically
valid:
∂ 2 γ xy∂ 2 ε xx ∂ ε yy
2

= + (11)
∂x ∂ y ∂ y 2 ∂x 2
That compatibility equation can be reformulated by the introduction of material law, here
using the plane stress material law in Eq. (4), which substituted into Eq. (11) yields

∂ 2 ⎛ 2(1+ ν ) ⎞ ∂ 2 ⎛ σ xx σ yy ⎞ ∂ 2 ⎛ σ yy σ xx ⎞
⋅ σ = ⎜ − ν ⋅ ⎟ + ⎜ − ν ⋅ (12)
∂x ∂ y ⎜⎝ E xy ⎟
⎠ ∂ y2 ⎝ E E ⎠ ∂x 2 ⎝ E E ⎟⎠

where E cancels and the new compatibility equation simplifies to

∂ 2 σ xy ∂2 ∂2
⋅ 2 ⋅(1+ ν ) = (σ − ν ⋅ σ ) + (σ − ν ⋅ σ xx ) (13)
∂x ∂ y ∂ y 2 xx yy
∂x 2 yy
As shown in Section 16 of Timoshenko’s book on the theory of elasticity Eq. (13) can be
further modified by introducing equilibrium (Timoshenko and Goodier 1969). To achieve
this the first step is to write the equilibrium equations in 2D, a straightforward
simplification of the 3D version:
∂σ xx ∂σ yx
+ + fx = 0 (14)
∂x ∂y

2D Elasticity Theory Updated May 22, 2019 Page 3

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

∂σ xy ∂σ yy
+ + fy = 0 (15)
∂x ∂y
Assuming the body forces are uniform and differentiating Eq. (14) with respect to x and
Eq. (15) with respect to y yields
∂ 2 σ xx ∂ σ yx
2

+ =0 (16)
∂x 2 ∂ y ∂x

∂ 2 σ xy ∂ 2 σ yy
+ =0 (17)
∂x ∂ y ∂ y2
Adding Eqs. (16) and (17), i.e., adding zero with zero yields
∂ 2 σ xy ∂ 2 σ yy ∂ 2 σ xx
2⋅ + + =0 (18)
∂x ∂ y ∂ y2 ∂x 2
This equilibrium equation is now solved for , which is substituted into Eq. (13) to obtain
the following new compatibility equation:

1 ⎛ ∂ σ yy ∂ 2 σ xx ⎞
2
∂2 ∂2
− ⋅⎜ + ⎟ ⋅ 2 ⋅(1+ ν ) = (σ − ν ⋅ σ ) + (σ − ν ⋅ σ xx ) (19)
2 ⎝ ∂ y2 ∂x 2 ⎠ ∂ y 2 xx yy
∂x 2 yy

Simplification yields the compatibility equation in terms of stresses

∂ 2 σ yy ∂ 2 σ xx ∂ 2 σ xx ∂ σ yy
2

− − = + (20)
∂ y2 ∂x 2 ∂ y2 ∂x 2
which can be written
⎛ ∂2 ∂2 ⎞
(
⎜⎝ ∂x 2 ∂ y 2 ⎟⎠ σ xx + σ yy = 0
+ ) (21)

This equation holds also for plane strain and has a non-zero right-hand side when the
body forces are not uniform (Timoshenko and Goodier 1969):
⎛ ∂2 ∂2 ⎞ 1 ⎛ ∂ fx ∂ f y ⎞
+ σ (
⎜⎝ ∂x 2 ∂ y 2 ⎟⎠ xx + σ yy
= − ) +
1− ν ⎜⎝ ∂x ∂ y ⎟⎠
(22)

One approach to solve 2D continuum problems is to seek stresses that imply equilibrium
and that satisfy the compatibility equations presented above, together with problem-
specific boundary conditions. A clever approach to achieve this is a concept called stress
functions.

Stress Functions
In many boundary value problems in structural mechanics it is possible to combine
equilibrium, material law, and kinematic compatibility equations into one governing

2D Elasticity Theory Updated May 22, 2019 Page 4

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

differentiation equation. This equation, together with problem-specific boundary

conditions on forces and displacements, are used to obtain solutions. In the 2D theory of
elasticity, and certain other problems, an additional helpful concept is the “stress
functions” introduced by George Biddell Airy (1801-1892) in an 1862 paper. The stress
function itself does not have physical meaning. Instead it is an auxiliary quantity, here
denoted j(x,y), from which stresses are derived:
∂2ϕ
σ xx = − fx ⋅ x (23)
∂ y2

∂2ϕ
σ yy = − fy ⋅ y (24)
∂x 2
∂2ϕ
τ xy ≡ σ xy = − (25)
∂x ∂ y
Notice that the double derivative of the stress function in one direction produces axial
stress in the perpendicular direction. That is perhaps the most physical interpretation that
is possible for the stress function. Next it is interesting to combine the stress function
with the equations for equilibrium, material law, and kinematic compatibility. Starting
with equilibrium, Eqs. (23)-(25) are substituted into Eqs. (14) and (15) yielding
⎛ ∂2ϕ ⎞ ⎛ ∂2ϕ ⎞
d ⎜ 2 − fx ⋅ x⎟ d ⎜ −
⎝ ∂y ⎠ ⎝ ∂x ∂ y ⎟⎠ ∂2ϕ ∂2ϕ
+ + fx = 2 − fx − + fx = 0 (26)
dx dy ∂ y ∂x ∂x ∂ y 2
⎛ ∂2ϕ ⎞ ⎛ ∂2ϕ ⎞
d⎜− d − f ⋅ y
⎝ ∂x ∂ y ⎟⎠ ⎜⎝ ∂x ⎟⎠
2 y
∂2ϕ ∂2ϕ
+ + fy = − 2 + 2 − fy + fy = 0 (27)
dx dy ∂x ∂ y ∂x ∂ y
which demonstrates that the equilibrium equations are automatically satisfied with the
stress function definition in Eqs. (23)-(25). Only material law and kinematic
compatibility remain and those have already been combined in the compatibility
equations presented earlier. Substitution of the stress function in Eqs. Eqs. (23)-(25) into
the compatibility equation in Eq. (21) and assuming uniform body forces yields
∂4ϕ ∂4ϕ ∂4ϕ
+ 2 ⋅ + =0 (28)
∂x 4 ∂ y 2 ∂x 2 ∂ y 4
This equation is the governing equation for 2D continuum problems. Solutions are found
by establishing stress functions that satisfy Eq. (28) together with problem-specific
boundary conditions. Once the stress function is determined the stresses are found from
Eqs. (23)-(25). However, many real in-plane members have shapes and boundary
conditions that render such analytical solutions unattainable. The numerical finite
element method provides and alternative in those circumstances.

2D Elasticity Theory Updated May 22, 2019 Page 5

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

2D Elastic Beams
In other documents on this website, the Euler-Bernoulli and Timoshenko beam theories
are described. Both those theories assume that plane sections remain plane and
perpendicular to the neutral axis. That is not an exact description of reality and this
document describes an alternative, based on the 2D theory of elasticity. In particular, a
cantilevered beam in the x-z-plane is considered in the following, as shown in Figure 2.
In another document on 2D elasticity theory, the 4th-order differential equation for the
stress function is derived. Written in terms of x and z it reads
∂4 F ∂4 F ∂4 F
+ + =0 (29)
∂x 4 ∂z 4 ∂x 2 ∂z 2

h x

P
L
Figure 2: Beam on elastic foundation.

The solution for specific problems is established by formulating a stress function that
satisfies the unique boundary conditions for each problem. As a demonstration, the
cantilever in Figure 2 is considered in this section. To simplify the mathematical
expressions, the fixed support is on the right-hand side and the load is applied on the left-
hand side. A solution for this problem, where the horizontal edges are stress-free and the
edge at x=0 has a shear stress resultant P, is obtained by combining the stress function for
pure shear and a stress function with a 3rd-order term (Timoshenko and Goodier 1969):
F(x, z) = C1 ⋅ x ⋅ z + C2 ⋅ x ⋅ z 3 (30)
This stress function yields the following stresses:
∂2 F
σ xx = = 6 ⋅C2 ⋅ x ⋅ z (31)
∂z 2
∂2 F
σ zz = 2 = 0 (32)
∂x
∂2 F
τ xz = − = −C1 − 3⋅C2 ⋅ z 2 (33)
∂x ∂z
Zero stress on the horizontal edges implies that

2D Elasticity Theory Updated May 22, 2019 Page 6

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

2
⎛ h⎞ 4
τ xz (z = ± 2 ) = −C1 − 3⋅C2 ⋅ ⎜ ⎟ = 0
h ⇒ C2 = − ⋅C (34)
⎝ 2⎠ 3h2 1
Shear stress resultant P at x=0 implies that
h h
2 2
2h 3P
∫
− h2
τ xz (x = 0) dz = ∫ −C
− h2
1
− 3⋅C2 ⋅ z 2 dz = − C = −P
3 1
⇒ C1 =
2h
(35)

which implies that

4 3P 2P
C2 = − 2
⋅ =− 3 (36)
3⋅ h 2h h
Substitution of C1 and C2 into Eq. (31) yields the axial stress
12P P
σ xx = 6 ⋅C2 ⋅ x ⋅ z = − 3
⋅x⋅z = − ⋅x⋅z (37)
h I
where I=h3/12 is introduced for this unit-width beam to echo the notation in elementary
beam theory. In fact, the axial stress is distributed exactly as in elementary beam theory.
Substitution of C1 and C2 into Eq. (33) yields the shear stress

3P ⎛ ⎛ 2z ⎞ ⎞
2
3P 6P 2
τ xz = −C1 − 3⋅C2 ⋅ z = − + 2
z =− ⎜ 1− ⎟ (38)
2h h3 2h ⎝ ⎜⎝ h ⎟⎠ ⎠

where it is observed again that the solution is exactly as in the elementary beam theory,
assuming that the load P is applied on the edge distributed according to the parabola in
Eq. (38). However, more information is available from this elasticity solution than beam
theory. In particular, displacements can be calculated, and these reveal that plane sections
do not remain plane during bending. To see this, i.e., to compute the displacements
associated with the stresses above, the general kinematics equations are invoked, which
read
∂u
ε xx = (39)
∂x
∂w
ε zz = (40)
∂z
∂u ∂w
γ xz = + (41)
∂z ∂x
The material law is also needed, which for plane stress (thin beam in the y-direction) read
1
ε xx = ⋅ (σ xx − ν ⋅ σ zz ) (42)
E
1
ε zz = ⋅ (σ zz − ν ⋅ σ xx ) (43)
E

2D Elasticity Theory Updated May 22, 2019 Page 7

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

τ xz
γ xz = (44)
G
where G=E(2(1+n)) is the shear modulus. The material law for plane strain (infinitely
thick beam in the y-direction) is obtained by replacing E by E/(1–n2) and n by n/(1–n)
(Timoshenko and Goodier 1969), which yields

ε xx =
1
E
(
⋅ (1− ν 2 ) ⋅ σ xx − ν ⋅(1+ ν ) ⋅ σ zz ) (45)

ε zz =
1
E
(
⋅ (1− ν 2 ) ⋅ σ zz − ν ⋅(1+ ν ) ⋅ σ xx ) (46)

τ xz
γ xz = (47)
G
For the plane stress case, the combination of the kinematics equations (39) to (41) with
the material law equations (42) to (44), and substitution of stresses from Eqs. (37) and
(38) together with szz=0 yields
∂u 1 σ P
ε xx = = ⋅ (σ xx − ν ⋅ σ zz ) = xx = − ⋅x⋅z (48)
∂x E E EI
∂w 1 1 P
ε zz = = ⋅ (σ zz − ν ⋅ σ xx ) = − ⋅ ν ⋅ σ xx = ⋅ν ⋅ x ⋅ z (49)
∂z E E EI

3P ⎛ ⎛ 2z ⎞ ⎞
2
∂u ∂w τ xz
γ xz = + = =− ⎜ 1− ⎟ (50)
∂z ∂x G 2Gh ⎝ ⎜⎝ h ⎟⎠ ⎠

Integration of Eqs. (48) and (49) yields the general displacement expressions
P Pzx 2
u(x, z) = − ∫ ⋅ x ⋅ z dx = − + hz (z) (51)
EI 2EI
P Pν xz 2
w(x, z) = ∫ ⋅ ν ⋅ x ⋅ z dz = + hx (x) (52)
EI 2EI
where hx(x) and hz(z) are functions that represent the integration constants. Substitution of
Eqs. (51) and (52) into Eq. (50) yields

3P ⎛ ⎛ 2z ⎞ ⎞
2
∂u ∂w Px 2 ∂hz (z) Pν z 2 ∂hx (x)
+ =− + + + =− ⎜ 1− ⎟ (53)
∂z ∂x 2EI ∂z 2EI ∂x 2Gh ⎝ ⎜⎝ h ⎟⎠ ⎠

By separating terms that depend on x and z, the following reorganized equation is

obtained:
Px 2 ∂hx (x) Pν z 2 ∂hz (z) 6Pz 2 3P
− + + + − =− (54)
!2EI
##"## ∂x$ ! 2EI
###" ∂z### 3
Gh$ !" 2Gh
$
F (x) G( z ) K

2D Elasticity Theory Updated May 22, 2019 Page 8

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

where F(x), G(z), and K are defined so that Eq. (53) can be written (Timoshenko and
Goodier 1969)
F(x) + G(z) = K (55)
Because the right-hand side is constant, F(x) and G(z) must also be constants, here named
C3 and C4:
Px 2 ∂hx (x) ∂hx (x) Px 2
F(x) = − + = C3 ⇒ = + C3 (56)
2EI ∂x ∂x 2EI
Pν z 2 ∂hz (z) 6Pz 2 ∂hz (z) 6Pz 2 Pν z 2
G(z) = + − = C4 ⇒ = − + C4 (57)
2EI ∂z Gh3 ∂z Gh3 2EI
Integration yields hx(x) and hz(z) expressed in terms of unknown constants:
⎛ Px 2 ⎞ Px 3
hx (x) = ∫ ⎜ + C3 ⎟ dx = + C3 x + C5 (58)
⎝ 2EI ⎠ 6EI

⎛ 6Pz 2 Pν z 2 ⎞ 2Pz 3 Pν z 3
hz (z) = ∫ ⎜ 3
− + C 4⎟
dz = − + C4 z + C6 (59)
⎝ Gh 2EI ⎠ Gh3 6EI
Substitution of the functions hx(x) and hz(z) in Eqs. (58) and (59) into the displacements
in Eqs. (51) and (52) yields
Pzx 2 2Pz 3 Pν z 3
u(x, z) = − + − + C4 z + C6 (60)
2EI Gh3 6EI
Pν xz 2 Px 3
w(x, z) = + + C3 x + C5 (61)
2EI 6EI
The constants C3, C4, C5, and C6 are determined from Eq. (55), which says that
3P
C3 + C4 = − (62)
2Gh
and from three boundary conditions that prevent the beam from displacing as a rigid
body. For the beam in Figure 2, it is natural to enforce zero displacements and zero
rotation at the point where x=L and z=0. According to Eqs. (60) and (61), zero
displacement implies that
u(L,0) = C6 = 0 (63)

PL3
w( L,0) = + C3 L + C5 = 0 (64)
6EI
Interestingly, the zero-rotation-requirement can be imposed in several ways because
plane sections do not remain plane and perpendicular to the neutral axis. One option is to
require that a vertical line remains vertical:

2D Elasticity Theory Updated May 22, 2019 Page 9

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

∂u( L,0) PL2 PL2

=− + C4 = 0 ⇒ C4 = (65)
∂z 2EI 2EI
Eq. (62) then yields
3P PL2
C3 = − − (66)
2Gh 2EI
Eq. (64) then yields
PL3 3PL PL3 PL3 3PL
C5 = − + + = + (67)
6EI 2Gh 2EI 3EI 2Gh
which means that all the constants C3, C4, C5, and C6 are determined and the final
displacement expressions are:
Pzx 2 2Pz 3 Pν z 3 PL2
u(x, z) = − + − + ⋅z (68)
2EI Gh3 6EI 2EI

Pν xz 2 Px 3 ⎛ 3P PL2 ⎞ PL3 3PL

w(x, z) = + +⎜− − x + + (69)
2EI 6EI ⎝ 2Gh 2EI ⎟⎠ 3EI 2Gh
Another option for enforcing zero rotation at the end is the classical requirement that a
horizontal line remains horizontal:
∂w( L,0) PL2 PL2
= + C3 = 0 ⇒ C3 = − (70)
∂x 2EI 2EI
Eq. (62) then yields
PL2 3P
C4 = − (71)
2EI 2Gh
Eq. (64) then yields
PL3 PL3 PL3
C5 = − + = (72)
6EI 2EI 3EI
which means that all the constants C3, C4, C5, and C6 are determined and the final
displacement expressions are:

Pzx 2 2Pz 3 Pν z 3 ⎛ PL2 3P ⎞

u(x, z) = − + − + − z (73)
2EI Gh3 6EI ⎜⎝ 2EI 2Gh ⎟⎠

Pν xz 2 Px 3 PL2 PL3
w(x, z) = + − x+ (74)
2EI 6EI 2EI 3EI
It is interesting to note that the tip-deflection of the cantilever equals the classical result
PL3/3EI for Eq. (74) while for the solution in Eq. (69) it is

2D Elasticity Theory Updated May 22, 2019 Page 10

Professor Terje Haukaas The University of British Columbia, Vancouver terje.civil.ubc.ca

PL3 3PL
w(x, z) = + (75)
3EI 2Gh
To observe the distortion of a plane cross-section during bending, the deformed shape of
an extreme-case beam (h=1m, L=1m, E=200,000N/mm2, n=0.3, P=10MN) is plotted in
Figure 3. The solid line shows the solution given by Eq. (68) and shows that the cross-
section remains vertical at x=L and z=0. However, for other z-values it has clearly
distorted from being a straight line. The dashed line shows the solution given by Eq. (73),
where the horizontal line at x=L and z=0 remains horizontal. Note that the angle between
the neutral axis and the cross-section at z=0 is obtained from Eq. (69):
∂w( L,0) 3P
=− (76)
∂x 2Gh

Figure 3: Cross-section distortion in the x-z-plane.

References
Timoshenko, S., and Goodier, J. N. (1969). Theory of elasticity. McGraw-Hill.

2D Elasticity Theory Updated May 22, 2019 Page 11