Lecture 9 - Plate Analysis
Lecture 9 - Plate Analysis
Plate analysis
Floor and deck slabs
Print version Lecture on Theory of Elasticity and Plasticity of Dr. D. Dinev, Department of Structural Mechanics, UACEG
9.1
Contents
1 2 3 4 5 6 7 Introduction Assumptions Field equations Equilibrium equations Principal values of the internal forces Boundary conditions 1 3 4 6 8 9
Analytical solutions 11 7.1 Naviers solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 7.2 L evys solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Numerical methods Engineering methods 14 17
9.2
8 9
Introduction
Introduction
Introduction
Introduction
Introduction
Introduction
Introduction
q t b a
Denition Plates are plane, 2-D structural components of which one dimension, called thickness t is much smaller than the other dimensions The plate loads are mainly transversal to the plane surface They are carried by internal bending and twisting moments and shear forces The plate edges can be simply supported, xed or elastically restrained
9.8
Assumptions
Assumptions Classical plate theory 1. 2. 3. 4. Note Applying the above assumptions we can reduce the 3-D problem to a 2-D plate bending problem This theory is known as a Kirchhoff-Love plate theory
9.9
t t 1 1 The plate is thin- t a, b and a , b = 10 50 The in-plane strains are small compared to the unity- xx , yy , xy The transverse normal strain is negligible- zz 0 The transverse shear stresses are negligible- xz , yz 0
Assumptions
Field equations
Field equations Displacements The assumption # 3 implies that w(x, y, z) = w0 (x, y) where w0 is the transverse displacement of the mid-plane (z = 0) Applying the assumption of the Kirchhoff hypothesis (plane section, normal to the midsurface before deection remains plane and normal to the deformed surface) gives u(x, y, z) = zx The assumption # 4 gives xz = yz = 0 Therefore x = w and y = w x y
9.11
Introduction
x, u q(x,y) z w0 x, u
z, w y, v
P z, w
w0 x
Displacements Plate kinematics
9.12
Field equations Plate kinematics Displacement eld w0 x w0 v(x, y, z) = z y w(x, y, z) = w0 (x, y) u(x, y, z) = z Strain-displacement eld w2 x2 w2 yy = z 2 y w2 yy = 2z x y xx = z
9.13
w2 x
1 0
0 0
1 2
xx z yy xy
Field equations
z y
M xy M yx My
Mx
M=
t/2
zdz =
t3 E 12
9.16
Field equations Internal forces Stress resultants (internal forces) 1 Mxx Myy = D Mxy 0
3
1 0
0 0
1 2
xx yy xy
Et where D = 12(1 2 ) is called exural rigidity (stiffness) of the plate The bending and twisting moments can be expressed in terms of displacements
Mxx = D
2w 2w + 2 x2 y 2w 2w + 2 x2 y 2w x y
9.17
Myy = D
Mxy = D(1 )
Equilibrium equations
Equilibrium equations
M xy dy dy M x dy Qx dy M y My + dy dx y Qy Qy + y dx M yx dx
M y dx
Qy dx
qdxdy M xy dx dy M xy + x M yx M yx + dy dx y dy dx
M x dx dy Mx + x Q Qx + x dx dy x
Equilibrium equations Cartesian coordinate system The moment equilibrium equations of a differential element lead to Mx Myx + x y My Mxy Qy = + y x Mxy = Myx Qx =
9.19
Equilibrium equations Cartesian coordinate system Using the above equilibrium relations we may obtain a single equation of the plate equilibrium in terms of the internal forces 2 Mxy 2 My 2 Mx +2 + = q x2 x y y2 Replacement of the moments-displacements relations gives the equilibrium equation in terms of the transversal displacement 4w 4w 4w q +2 2 2 + 4 = x4 x y y D
9.20
Equilibrium equations
Joseph-Louis Lagrange (1736-1813) Cartesian coordinate system Sophie Germain- Lagrange equation 4 w(x, y) = q(x, y) D
9.22
Equilibrium equations Shear forces The shear forces also can be expressed in terms of the displacements Qx = D Qy = D x y 2w 2w + 2 x2 y 2w 2w + 2 x2 y
9.23
Principal bending moments Consider the internal forces acting on plate with arbitrary section cut Applying the equilibrium equation z = 0 we have Qn = Qx cos + Qy sin
9.24
Principal values of the internal forces Principal bending moments The moment equations (Ms = 0 and Mn = 0) gives Mnn = Mxx cos2 + 2Mxy sin cos + Myy sin2 Mns = (Myy Mxx ) sin cos + Mxy (cos2 sin2 ) The innitesimal parts of Qx , Qy and q are neglected
9.25
Principal values of the internal forces Principal bending moments The extremum condition
Mnn
9.26
Boundary conditions
q t b a
Boundary conditions
Loading and supports An exact solution of the governing plate equations must simultaneously satisfy the differential equations and BCs of any given plate bending problem Since the 8-th-order differential equation require two boundary conditions at each plate edge
9.27
Boundary conditions
Boundary conditions
Natural (force) BCs Mx = 0 Vx = 0 Modied shear force (Kirchhoff equivalent force) Mxy y Mxy Vy = Qy + x Vx = Qx +
9.29
Boundary conditions
Boundary conditions
Boundary conditions
Corner forces 10
When we have a corner at the plate boundary the twisting moments jump from +Mxy to Mxy The jump of the twisting moment is called corner force Rc Rc = 2Mxy
9.32
Boundary conditions
Corner forces This effect appears at plates with corners with simply supported edges Question What happens when the edges are xed or free?
9.33
7
7.1
Analytical solutions
Naviers solution
x z y
Analytical solution
b a
Naviers solution- double Fourier series Naviers solution- solution by double trigonometric series Rectangular plate Boundary conditions w=0 w=0 Mx = 0 My = 0 at x = 0, x = a at at y = 0, y = b y = 0, y = b
9.34
at x = 0, x = a
11
Analytical solution
Analytical solution Naviers solution- double Fourier series Suppose that the solution is
w(x, y) =
m=1 n=1
wmn sin
n m x sin y a b
The above solution satises the BCs The coefcient of expansion wmn is unknown The transversal load also can be expanded into double series
q(x, y) =
m=1 n=1
qmn sin
m n x sin y a b
9.36
Analytical solution Naviers solution- double Fourier series Substitution of the above relations into the equilibrium equation gives wmn m4 4 m2 n2 4 n4 4 +2 2 2 + 4 4 a a b b qmn m n = sin x sin y D a b wmn = D 4 The nal solution is w(x, y) = 1 D 4 m=1 n=1 qmn
m2 a2 n2 b2 2
sin
m n x sin y a b
Hence qmn
m2 a2 n + b2
2
sin
m n x sin y a b
9.37
Analytical solution Naviers solution- double Fourier series When the transverse displacements is obtained we may calculate the internal forces Internal moments
Mx = 2 D My = 2 D
m=1 n=1 m=1 n=1
m a
+
2
n b n b
wmn sin
2
m n x sin y a b m n x sin y a b
m a
wmn sin
Mxy = 2 D(1 )
9.38
12
+
2
n b n b
+ (1 ) + (1 )
n b m a
mwmn cos
2
3D m a b m=1 n=1
nwmn sin
7.2
L evys solution
Analytical solution L evys solution- single Fourier series L evys solution- solution by single trigonometric series Applicable to rectangular plates, simply supported at two opposite edges The solution of the equilibrium equation is given by w(x, y) = wh (x, y) + w p (x) The particular solution is
w p (x) = wm sin
i=1
m x a
m x a
9.40
Analytical solution L evys solution- single Fourier series The replacement of the above expressions into the equilibrium equation gives a differential equation for ym m4 4 m2 2 fm (y) 2 2 fm (y) + fm (y) = 0 a4 a The solution of the above equation is fm (y) =Am cosh m m m y + Bm y sinh y a a a m m m +Cm sinh y + Dm y cosh y a a a
9.41
where Am , Bm , Cm = 0 and Dm = 0 are constants and can be determined from the BCs.
Analytical solution L evys solution- single Fourier series The nal solution is
w(x, y) = wm sin
i=1
m x a m m m m y + Bm y sinh y sin x a a a a
9.42
+ Am cosh
i=1
13
Numerical methods
Numerical methods Approximate solutions The universal analytical solution of the governing plate bending equations for complex domain geometry, BCs and loading is not possible to nd The engineering practice needs to use approximate solutions to solve the above mentioned problems The approximate solution are based on the energy and variational methods of structural mechanics Ritz method Galerkin method Kantorovich method Numerical methods Finite differences method Gridwork method Finite elements method Finite strip method
9.43
Numerical methods
2
Example- FE analysis
Determine the displacements and the moments resultants for the given problem E = 20000000kPa, = 1/4, t = 0.1m
9.44
Numerical methods
14
4.1
6 6 21=q m5
m/Nk 5=q
5=q
Numerical methods
Numerical methods
Numerical methods
Numerical methods
15
Numerical methods
Numerical methods
Numerical methods
16
Numerical methods
Numerical methods
Engineering methods
Engineering methods
17
The engineering approach known as Marcus method (1924) The plate is considered as an elastic web consisting of plate strips located at mid-spans of the individual panels The application of this method is limited to Uniform load The size difference of the neighboring panels less than 50% 1 The Poissons ratio is = 6
9.55
Engineering methods Elastic web analogy Mid-span deections of the webs 1 qx 4 x 384 Dx 2 qy 4 y (wy ) y /2 = 384 Dy (wx ) = x /2 Because of Dx = Dy = D and (wx )
x /2
= (wy )
y /2
q = qx + qy
9.56
Engineering methods Elastic web analogy We may obtain the directional loads qx = 2 4 y q 4 +2 4 y x
Engineering methods
yL
1 esaC
xL
18
Engineering methods
Engineering methods
Engineering methods
yL
yL
yL
3 esaC
2 esaC
4 esaC
xL
xL
xL
19
Engineering methods
Engineering methods
Engineering methods
yL
yL
6 esaC
5 esaC
xL xL
The fundamental difference between the grillage and the plate is the presence of shear forces between individual strips which produce a torsional resistance an reduce the deections
9.64
Engineering methods Elastic web analogy The approximated maximum span moments in plate are Mx = Mx 1 5 6
2 x 2 y 2 y 2 x
Mx 0 Mx My 0 My
5 My = My 1 6
q 2 x , 8
0 My =
q 2 y 8
9.65
Engineering methods Elastic web analogy The edge moments are calculated as a strip supported with the same type of supports as a plate and loaded with directional load qx or qy 1 qx 2 12 x 1 supp My = qy 2 8 y
supp Mx =
9.66
Engineering methods
Elastic web analogy When the support moments of the neighboring panels does not match the bending moments can be averaged or calculated as support moments in a continuous beam
9.67
21
Engineering methods
2
Engineering methods
22
4.1
6 6 21=q m5
m/Nk 5=q
5=q