Derivation of Governing Equations in terms of

displacements for the static bending and buckling

of a nanobeam considering shear deformation
based on the NSGT

March 25, 2022

1 DERIVATION
1.1 Governing Equations
The displacement fields are given by,

 
dw(x) dw(x)
ux (x, z) = u(x) − z + R(z) − φ(x) (1)
dx dx
uy (x, z) = 0 (2)
uz (x, z) = w(x) (3)

where u, w and φ are the axial displacement, transverse displacement

and rotation angle of the cross section about y-axis, respectively. R(z)
is a function of z and satisfies the zero transverse shear stress and strain
condition at the top and bottom surfaces of the beam without any shear
correction factors.
The non-zero strain components can be obtained by differentiating equa-
tion (1),(2) and (3),

1
 ∂ux du dφ d2 w
εxx = = −R − (z − R) 2 (4)
∂x
 dx dx
  dx 
1 ∂ux ∂uz 1 dR dw
εxz = + = −φ (5)
2 ∂z ∂x 2 dz dx

Using the Principle of virtual work we know that,

δU − δW =0 (6)

The Virtual Stain Energy is given by,

Z  
(1) (1)
δU = σxx δεxx + 2σxz δεxz + σxx ∇δεxx + 2σxz ∇δεxz dV
Z V Z Z Z
(1) (1) (1)
= (σxx δεxx + 2σxz δεxz ) dV + σxx δεxx dA − ∇σxx δεxx dV + σxz δεxz dA
Z V A V A
(1)
− ∇σxz δεxz dV
V
Z   Z   Z Z L
(1) (1) (1) (1)
= σxx − ∇σxx δεxx dV + 2 σxz − ∇σxz δεxz dV + σxx δεxx dA +2 σxz δεxz dA
V V A A 0
Z Z Z Z L
(1) (1)
δU = (txx ) δεxx dV + 2 (txz ) δεxz dV + σxx δεxx dA + 2 σxz δεxz dA (7)
V V A A 0

Substitute the equations (4) and (5) into equation (7),

d2 δw
Z   Z   
dδu dδφ 1 dR dδw
δU = txx −R − (z − R) dV + 2 txz − δφ dV
V dx dx dx2 V 2 dz dx
d2 δw
Z   Z   
(1) dδu dδφ (1) 1 dR dδw
+ σxx −R − (z − R) dA + 2 σxz − δφ dA
A dx dx dx2 A 2 dz dx
Z L
d2 δw
 
dδu dδφ dδw
δU = N −M −P + Q − δφ dx
0 dx dx dx2 dx
 2
 L
(1) dδu (1) dδφ (1) d δw (1) dδw
+ N −M −P +Q − δφ (8)
dx dx dx2 dx 0

2
 where the stress resultants are defined as,

Z Z
dR
{N, M, P } = txx {1, R, (z − R)}dA, Q = txz dA (9)
dz
ZA A
Z
(1) dR
{N (1) , M (1) , P (1) } = (1)
σxx {1, R, (z − R)}dA, Q(1) = σxz dA (10)
A A dz

The Virtual Work done is given by,

Z L 
dw dδw
δW = f δu + qδw + Nm dx (11)
0 dx dx

L
d2 δw
Z  
dδu dδφ dδw
δU − δW = 0 = N −M −P +Q − δφ dx
0 dx dx dx2 dx
L
d2 δw
 
dδu dδφ (1) dδw
+ N (1) − M (1) − P (1) + Q − δφ
dx dx dx2 dx 0
Z L   
dw dδw
− f δu + qδw + Nm dx
0 dx dx
Z L  Z L  Z L 
dN dM dδw dP dδw
=N δU − δudx − M δφ + δφdx − P + dx +
0 dx 0 dx dx 0 dx dx
Z L  Z L Z L Z 
dQ dQ
Qδw − δwdx − Q (δφdx) + (δφdx) dx
0 dx 0 0 dx
Z Z  Z 2 
dw d w
− qδwdx − f δudx − Nm δw − δwdx
dx dx2
 2
 L
(1) dδu (1) dδφ (1) d δw (1) dδw
N −M −P +Q − δφ
dx dx dx2 dx 0

Again doing partial integration of

Z L 
dP dδw
dx
0 dx dx
we get,

3
 Z L  Z L 
dN dM dδw dP
δU − δW = 0 =N δU − δudx − M δφ + δφdx − P + δw−
0 dx 0 dx dx dx
Z 2 Z L  Z L Z L Z 
d P dQ dQ
δwdx + Qδw − δwdx − Q (δφdx) + (δφdx) dx
dx2 0 dx 0 0 dx
Z Z  Z 2 
dw d w
− qδwdx − f δudx − Nm δw − δwdx
dx dx2
 2
 L
(1) dδu (1) dδφ (1) d δw (1) dδw
N −M −P +Q − δφ (12)
dx dx dx2 dx 0

Applying Principal of fundamental lemma,

 
dN
δu : − −f =0
dx
 
dM
δφ : −Q =0
dx
d2 w d2 P
 
dQ
δw : − − q + Nm 2 − =0
dx dx dx2

Rearranging the terms we get,

 
dN
δu : +f =0 (13)
dx
 
dM
δφ : −Q =0 (14)
dx
d2 P d2 w
 
dQ
δw : +q+ − Nm 2 = 0 (15)
dx dx2 dx

The simplified equation for higher-order nonlocal strain gradient theory

is given by,

1 − (ea)2 ∇2 txx = E 1 − l2 ∇2 εxx

 

(16)

4
 Substitute the equations (4) and (5) into equation (16),

d2 w
 
 2 2
 2 2

du dφ
1 − (ea) ∇ txx = E 1 − l ∇ −R − (z − R) 2 (17)
dx dx dx
  

dR dw
1 − (ea)2 ∇2 txz = G 1 − l2 ∇2
 
−φ (18)
dz dx

where E and G are the Young’s modulus and shear modulus, respectively.
With the help of Eqs. (17) and (18), the stress resultants defined in Eqs.
(9) can be obtained as, Z
N= txx dA
Z
2 2 2 2



N 1 − (ea) ∇ = 1 − (ea) ∇ txx dA

d2 w
Z  

du dφ
N 1 − (ea)2 ∇2 = E 1 − l 2 ∇2


−R − (z − R) 2 dA
dx dx dx
For Timoshenko beam Theory,

R=z
Z Z +h/2
du dφ
N 1 − (ea)2 ∇2 = E 1 − l2 ∇2



dA − z dz
dx −h/2 dx

du
N 1 − (ea)2 ∇2 = Axx 1 − l2 ∇2


dx
Similarly we get the other equations as,
2 2
 
2d N 2 d du
N − (ea) 2
= Axx 1 − l 2
(19)
dx dx dx
2 2 d2 w
  
2d M 2 d dφ
M − (ea) = 1−l −Bxx − Cxx 2 (20)
dx2 dx2 dx dx
2 2 2
  
2d P 2 d dφ d w
P − (ea) = 1−l −Cxx − Dxx 2 (21)
dx2 dx2 dx dx
2Q 2
   
d d dw
Q − (ea)2 2 = κExz 1 − l2 2 −φ (22)
dx dx dx

5
 where κ is the shear correction factor. For Timoshenko beam the-
ory, κ = 5/6. For higher-order beam theory, κ = 1. The coefficients
Axx , Bxx , Cxx , Dxx and Exz are given by,
Z
E 1, R2 dA

{Axx , Bxx } = (23)
ZA
{Cxx , Dxx } = E(z − R){R, z − R}dA (24)
A
dR 2
Z  
EXz = G dA (25)
A dz

Thus, the governing equations in terms of displacements for the static

bending and buckling of a nanobeam considering shear deformation based on
the NSGT can be obtained by inserting Eqs. (19)-(22) into Eqs. (13)-(15)
as,

2
  2  2

2 d d u 2 d
Axx 1 − l + 1 − (ea) f =0 (26)
dx2 dx2 dx2
d2 d2 φ d3 w d2
       
dw
1 − l2 2 −Bxx 2 − Cxx 3 − Exz 1 − l2 2 − φ = 0 (27)
dx dx dx dx dx
2 d3 φ d4 w 2
     2 
2 d 2 d d w dφ
1−l −Cxx 3 − Dxx 4 + Exz 1 − l −
dx2 dx dx dx2 dx2 dx
2 2
  
d d w
+ 1 − (ea)2 2 q − Nm 2 = 0 (28)
dx dx

In the special case of Euler-Bernoulli beam, R(z) = 0. One can obtain

that,
Axx = EA, Dxx = EI, Bxx = Cxx = Exz = 0 (29)

where A is the cross section area and I is the second moment. Substituting
Eq. (29) into Eq. (28), the governing equation for Euler-Bernoulli beam
based on the NSGT can be given by,
2 ∂4w 2 d2 w
    
2 ∂ 2 ∂
1−l EI 4 + 1 − (ea) −q + Nm 2 = 0 (30)
∂x2 ∂x ∂x2 dx

6
It should be noted that when the nonlocal parameter ea is taken to be zero,
Eqs. (26) - (28) can be reduced to the governing equations based on the
strain gradient theory. When setting the material length scale parameter
l equal to zero, the nonlocal beam model will be recovered. If both the
nonlocal parameter and material length scale parameter are equal to zero,
Eqs. (26) - (28) will be turned into the governing equations of classical
continuum theory.

1.2 Analytical Solutions for Buckling Problems

In present work, the Navier’s method is utilized to obtain the analytical so-
lutions for bending and buckling problems of simply supported nanobeams.
Thus, the generalized expansions of displacements (φ, w) are assumed as,

∞
X
φ(x) = Φn cos (βn x) (31)
n=1
X∞
w(x) = Wn sin (βn x) (32)
n=1

where βn = nπ/L, Φn and Wn are undetermined Fourier coefficients.

For buckling analysis, we set the external transverse load q = 0. Substi-
tuting Eqs. (31) and (32) into Eqs. (27) and (28), the two coupled equations
can be written in a matrix form as following,

d2 
 
2 2 3
  
1−l B xx Φ n βn Cos (βn x) + C xx W n β n Cos (β n x)
dx2
2
 
2 d
− Exz 1 − l [Wn βn Cos (βn x) − Φn Cos (βn x)] = 0
dx2

Bxx Φn βn2 Cos (βn x) + l2 Bxx Φn βn4 Cos (βn x) +

 

Cxx Wn βn3 Cos (βn x) + l2 Cxx Wn βn5 Cos (βn x) −

 

Exz Wn βn Cos (βn x) + l2 Exz Wn βn3 Cos (βn x) −

 

Exz Φn Cos (βn x) + l2 βn2 Exz Φn Cos (βn x) = 0

 

7
 Bxx Φn βn2 1 + l2 βn2 + Cxx Wn βn3 1 + l2 βn2 −

Exz Wn βn 1 + l2 βn2 + Exz Φn 1 + l2 βn2 = 0

Φn Bxx βn2 ξ + Exz ξ + Wn Cxx βn3 ξ − Exz βn ξ = 0

2
 
2 d
−Cxx Sin (βn x) βn3 Φn − Dxx Wn Sin (βn x) βn4 +
 
1−l 2
dx
2
 
2 d 2


Exz 1 − l −W n sin (β n x) β n + Φ n sin (β nx ) β n +
dx2
d2
 
1 − (ea)2 2 q + Nn Wn Sin (βn x) βn2 = 0


dx

−Cxx Sin (βn x) βn 3 Φn − l2 Cxx sin (βn x) βn5 Φn

 

−Dxx Wn sin (βn x) βn4 − l2 Dxx Wn sin (βn x) βn 6

 

−Exz Wn Sin (βn x) βn2 − Exz Wn Sin (βn x) βn4 l2 +

 

Φn Exz Sin (βn x) βn + Φn Exz Sin (βnx ) βn3 l2 +

 

Nm Wn Sin (βn x) βn2 + Nm Wn Sin (βn x) βn4 (ea)2 = 0

 

Cxx Φn βn3 1 + l2 βn2 + Dxx Wn βn4 1 + l2 βn2 +

Exz Wn βn2 1 + l2 βn2 − Exz Φn βn 1 + l2 βn2 +

Nm Wn βn2 1 + (ea)2 βn2 = 0

Φn Cxx βn3 ξ − Exz βn ξ + Wn Dxx βn4 ξ + Exz βn2 ξ −

Nm Wn βn2 λ = 0

8
     
s11 s12 Φn 0
= (33)
s12 s22 − Nm βn2 λ Wn 0

where,

s11 = Bxx βn2 ξ + Exz ξ, s12 = Cxx βn3 ξ − Exz βn ξ

s22 = Dxx βn4 ξ + Exz βn2 ξ, ξ = 1 + l2 βn2 , λ = 1 + (ea)2 βn2 (34)

The non-trivial solution for the buckling problem can be obtained by solving
the characteristic equation.
S11 S22 − Nm βn2 λ − S122


=0
S11 S22 − S11 Nm βn2 λ − S12
2
=0

The critical buckling load for Timoshenko and higher-order beam models is
given by,
2
S11 S22 − S12
cr
Nm = (35)
βn2 λS11

For Euler-Bernoulli beam, substituting Eq. (32) into Eq. (30), the critical
buckling load can be obtained as,
EIWn βn4 sin (βn x) + l2 EIWn βn6 sin (βn x) − Nm Wn βn2 Sin (βn x) − (ea)2 Nm Wn βn4 sin (βn x) = 0
2 2
   
2 ∂ 4 2 ∂
−q − Nm Wn βn2 sin (βn x) = 0


1−l 2
EIWn βn sin (βn x) + 1 − (ea) 2
∂x ∂x
q=0
EIWn βn4 sin (βn x) 1 + l2 β 2 − Nm Wn βn2 sin (βn x) 1 + (ea)2 βn2 = 0

EIβn4 1 + l2 βn2 − Nm βn2 1 + (ea)2 βn2 = 0

EIβn2 ξ
Nm = (36)
λ