Topics in Ship Structural Design: Lecture 5 Buckling and Ultimate Strength of Plates

Topics in Ship Structural Design
(Hull Buckling and Ultimate Strength)
Lecture 5 Buckling and

Ultimate Strength of Plates
Reference : Ship Structural Design Ch.12
NAOE
Jang, Beom Seon
OPen INteractive Structural Lab

Failure of MSC Napoli Container ship – DNV Report by Olav Nortun
Facts about MSC Napoli
 One of the world’s largest container ships when built (1991)
 Built to BV Class and changed to DNV 2002
 Last renewal survey carried out in 2004 in Singapore
 Built 1991
 Length over all 275.66 m
 Breadth 37.13 m
 Draught 13.50 m
 Gross tonnage 53,409 GRT
 Capacity 4419 TEU
Slide 2 OPen INteractive Structural Lab

Accident January 2007 – MSC Napoli
 Ship left Antwerp 17 January 2007
heading for Sines in Portugal
 18 January - water ingress in engine
room reported
 All 26 crew members safely rescued
 Ship beached in Lyme Bay near
Branscombe, UK on 19 January
2007

Accident January 2007 – MSC Napoli
 The vessels was split into two in
July 2007
 Forward part was towed to
Belfast for recycling

DNV Rule for Classification of Ships Part 3 Chapter 1, Section 13
B200 Plate panel in uni-axial compression
 Unstiffened Plate (Plating between stiffeners)

 Elastic and Inelastic Buckling
 Post-Buckling and Ultimate strength
Classification Rule
Ideal elastic buckling stress For plating with longitudinal stiffeners (in
2
direction of compression stress): k=4
 t 
 el  0.9kE   ( N/mm )
2
 1000 s 
Johnson-Ostenfeld plasticity correction formula

12.1 Elastic Plates Subjected to Uniaxial Compression
Buckling of a Wide Column
 The plate is acting more as a wide column than as a plate. The product EI is
replaced by the plate flexural rigidity D.
 2D  2E
2
 2 Db t
Pcr   cr  2   
a2 a t 12(1  v 2 )  a 
 The thickness / length ratio plays the same role as the slenderness ratio for
columns.
 The width b plays no part, no support along the unloaded edge → It is
inefficient to use
Simple support along the loaded edges
With no support along the unloaded edges
Buckling of wide column

6 OPen INteractive Structural Lab
9.2 Combined Bending and Membrane Stresses-Elastic Range
Large-Deflection Plate Theory by von Karman
 Small-Deflection Plate Theory
4w 4w 4w p
2 2 2  4 
x 4 x y y D
 Large-Deflection Plate Theory

4w 4w 4w 1  2w 2w 2w 
 2 2 2  4   p  N x 2  2 N xy  N y 2 
x 4 x y y D x xy y 
1 2w 2w 2w 

 w   p  N x 2  2 N xy
4
 N y 2 
D x xy y 

Buckling of a Simply Supported Plate
 From large-deflection plate theory
 at  2 w
N x   at , p  N y  N xy  0  w
4
D x 2
 Since the edges are simply supported, the
deflected shape can be expressed in the form:
m x n y
w   wmn  Cmn sin sin
m n m n a b
Buckled shape of long plate
which satisfies both the boundary conditions and
the general biharmonic equation.
 Plate is assumed to be free to move inward under the action of the in-plane
compression. → The strain energy of deformation is due to bending only
D a b  2 w  2 w 
2
  2 w  2 w   2 w  2  
U =    2  2  2(1  v)  2     dxdy
2 0 0  x y   x  y 2
 x y   

2
 4 ab 2 m
2
n2 
U D Cmn  2  2 
8 m n a b 

Reference
Strain Energy Density for plane stress (σz=0)
dxεx
dzεz dz dx
dyεy
dz
dy
Load applied Elongation in V 1

on dydz area x-direction du2   ( x dxdz)( xy dy)
2 2
1 1 1
du1  ( x dydz)(dx x )  ( y dxdz)(dy y )   x xy dxdydz
2 2 2
1
 ( x x   y y )dxdydz dx
2
dy thickness t
1
 du  ( x x   y y   x xy )dxdydz
2
Reference
Strain Energy for plane stress (σz=0)
 In Chapter 9 (Lecture 03), Plate bending (Derivation of Plate
Bending Equation), the followings are derived
E  2w 2w   2w 
x  ( z ) 2  v 2   x  ( z ) 2 
1 v 2
 x y   x 
E  2w 2w   2w 

y  ( z ) 2  v 2   y  ( z ) 2 
1 v2  y x   y 
2w  u  2w 
  2Gz = +   (2 z ) 
xy x y  xy 
w
u  z
x
w
v  z
y
a b t/2 1
U   du    t / 2 2 ( x x   y y   x xy )dzdxdy
0 0
Reference
Strain Energy Density for plane stress (σz=0)
1 a b t / 2 E   2 w  2 w   2 w  2 E
U   du     2 
 2  v 2  2 ( z ) G
2 0 0  t / 2 (1  v )  x y  x  2(1  v)
E  2w  2 w   2 w  2 2  w 
2
 2 
 2  v 2  2 ( z )  4Gz  dzdydx
(1  v )  y x  y   xy 
Et 3   2 w    2 w   2 w  2 w  4
2 2
E (1  v) 3  w 
2
a b
 0 12(1  v 2 )  x 2    y 2   2v x 2 y 2   12 2(1  v)(1  v) t  xy dydx
0
 
2 2 2
D a b  2w   2w  2w 2w  2w 
    2    2   2v 2  2(1  v)  dydx
2 0 0  x   y  x y 2  xy 
2 2
D a b  2w 2w   2w 2w 2w 
U     2  2   2(1  v) 2   dydx
 x y   x y xy 
2
2 0 0

Reference
Work done for plane stress
For unit-width strip in Section 9.2 dx
   
2  b 1  w 
2 w
w
 x    1     1dx     dx
b w
w dx
0  x   0 2 x x
    
  w  2 
1    dx
1  w 
2
a b a b   x  
W   N x x dxdy   0 a 2  x  dxdy
 t  
0 0 0
a
(1  a )  1  for small a
2

 Likewise, the work done by the in-plane compressive stress is
 a t a b  w   4b a t
2
W 
2 0 0  x  
 dxdy W
8a m n
 2
Cmn m2
 Because of W=U, and hence, 2 m

2
n2 
2 2
 a D  Cmn  2
 
a b2 
a  m n
t  m 2Cmn
2
m n
c  c  ...  cn
min(c1 / d1 , c2 / d 2 ,..., cn / d n )  1 2  max(c1 / d1 , c2 / d 2 ,..., cn / d n )
d1  d 2  ...  d n
 The minimum value of σa is given by taking only one term, say Cmn,
2
 a D  m2 n2 
2 2
( a )cr  2  2  2
tm  a b 
where m and n indicate the number of half-waves in each direction in the
buckled shape.
 When n=1, σa gives the smallest value. Hence the plate will buckle into only
one half-wave transversely.
2
 D 1 a 
2 2
( a )cr  2  m    
a t  m  b  
 A buckling coefficient k is generally used. It depends on the type of
2
boundary support.  2D  mb a 
( a )cr  k 2 k   
bt  a mb 
 For design applications, in which the plate thickness is to be determined, it
is usually written like this:
t
2
 2k Et 3
( a )cr  KE   K D
b 12(1  v 2
) 12(1- 2 )
Q: Which critical stress will be higher?, which stiffener

arrangement is better against in-plane compression?

 For long simply supported plates it is  2D

2
 mb a 
usually assumed that k=4. ( a )cr  k k   
b 2t  a mb 
 Assuming v=0.3
Homework #1 Plot this curve
 2D
( a )cr  4
b 2t
2
t
( a )cr  3.62 E  
b
Classification Rule
2
 t 
 el  0.9kE   ( N/mm )
2
 1000 s 
k=4, s=b (m)15
Buckled shape of long plate

a/b=1, m=1 a/b=2, m=2
a/b=3, m=3

 For a wide plate, in which the aspect ratio(a/b) is less than 1.0, m=1
2
 D a 
2 2
( a )cr  2 1     b
a t   b  
 For a general "wide plate“, in terms of a a

because a<b
2
b a
2
 D 2
a
( a )cr  k k   k k   
a 2t b a b
 For design purposes it may be written as:

2
  a 2 
2
t
( a )cr  KE   2  2 Et 3
K 2 
1    D
a 12(1  v )   b   12(1  v 2 )
For v=0.30
2
  a 2 
K  0.905 1    
  b  

 Longitudinal stiffeners: a>>b(=s), k=4
4 2 D
( a )cr  2
st
 Transverse stiffeners: a<<b, a=s, b=B,

m=1
2
 D  s  
2 2
( a )cr  2 1    
s t   B  
<<1
 Longitudinally stiffened plating have the great advantage over

transversely stiffened plating in ship structures, and the former is used
wherever possible.

Reproducing the event in a computer model
 Direct wave load calculations
 Linear strength analysis
 Non-linear strength analysis
 Load and strength comparisons
 Simulation of crack propagation

Most severe wave for engine room area
Hull forces: Shear force and moment
Aft ship out of water
Wave crest around midship
Vertical ”g” force

Failure of MSC Napoli Container ship
Structural arrangement in Engine room zone

Not sufficient buckling capacity
 The buckling capacity might not have been checked sufficiently

when the ship was built
 Potentially insufficient buckling strength in the engine room
bulkhead

Four stages of progressive collapse Outer shell

Four stages of progressive collapse Inner structure

Four stages of progressive collapse Inner structure

Alternative correcting actions
 The likelihood of reoccurrence is
very low:
 Damage statistics are very
good
 Little likelihood of such a harsh
sea state
 The ship’s strength was below
the strength of similar ships
 Maybe not all ships checked in
this area
 However – the consequences
are major
 Increase buckling strength

 Minor modifications – small
amount of steel to be added
 Aft of the engine room
bulkhead
 Can be done while in service

12.2 Other Boundary Conditions
Solutions for Some Principal Cases
Q: Which edge is more effective to in-

plane buckling? Loaded edge or unloaded 2
t
edge? ( a )cr  KE  
b
A, B fixed
B simply supported
A : Unloaded edge A fixed
B: Loaded edge
B fixed
A simply supported
 When unloaded edge (A) is replaced A & B simply
by simply supported, the critical supported
buckling stress drops more than when
loaded edge (B) is by simply supported.
Buckling coefficient k in the design formula

for flat plates in uniaxial compress
In general, b≈ 800mm,
Solutions for Some Principal Cases a≈3300mm, a/b≈3~4
 2D Unloaded edge : clamped

( a )cr  k Loaded edge : clamped
b 2t
Unloaded edge : clamped
Loaded edges
clamped Loaded edge : simply supported
Loaded edges
Simply supported Unloaded edge : pinned and clamped
A1, A2 clamped Loaded edge : clamped
Unloaded edge : pinned and clamped
A1 pinned Loaded edge : simply supported
A2 clamped
Unloaded edge : simply supported
Loaded edge : clamped
A1, A2 pinned
Unloaded edge : simply supported
Loaded edge : simply supported
Unloaded edge : free & clamped

Loaded edge : clamped
A1 free
A2 clamped Unloaded edge : free & clamped
A1 free
A2 pinned Unloaded edge : free & pinned
A1 A2 free Loaded edge : clamped
Unloaded edge : free & pinned
Buckling stress coefficient k for flat plates Unloaded edge : free
in uniaxial compression Loaded edge : simply supported
Solutions for Some Principal Cases
2
 D
2 t
( a )cr  k ( a )cr  KE  
b 2t b
Loaded edges
clamped
Loaded edges
Simply supported
A1, A2 clamped A, B fixed
A1 pinned
A2 clamped B simply supported
A fixed
A1, A2 pinned
B fixed
A simply supported
A & B simply
supported
A1 free
A2 clamped
A1 free
A2 pinned
A1 A2 free
Buckling stress coefficient k for flat plates Buckling coefficient k in the design formula
in uniaxial compression for flat plates in uniaxial compress
Clamped Edges
 For in-plane loads, as in the case of lateral loads, it is not possible to obtain
finite expressions for the solution of clamped plates.
 Numerical solutions by Faxen, Maulbetsch, and Levy.
Buckling coefficient k for clamped plates
under uniaxial compression

Unloaded Edges Rotationally Restrained
 Lundquist and Stowell have

investigated the case in which
the support along the unloaded
Loaded edges
edges is intermediate between clamped
simply supported and clamped.
 The degree of rotational restraint

is specified in terms of a
coefficient of restraint, defined as Loaded edges
Simply
supported
b
  Cy
D
 Cy : rotational stiffness of the

supporting structure along the
unloaded edge
Buckling coefficient k for plates with loaded
edges simply supported and longitudinal edges
31 rotationally
OPen restrained
INteractive Structural Lab
Loaded Edges Rotationally Restrained
 The important boundary conditions are those along the longer edges of the
plate. Thus, for short wide plates the edge restraint along the loaded edges
becomes significant.
 Similar to end conditions in a column, by using an effective length ae:
2
 D   ae   for clamped ends ae = 1/2a
2 2
( a )cr  2 1    
ae t   b  
for one end simply supported and
the other clamped ae = 0.707a
 Using a coefficient of restraint ζ :
a
  Cx
D
Cx : rotational stiffness of the supporting structure along the unloaded edge
 The solution to this case is obtained from
 k1  2  k2
K1 tan  K 2 tan 
( K1  K 22 )  0
2 2 
in which K1 and K2 are related to the buckling coefficient k.

K1,2  ( k  k  4)
2
Buckling coefficient 𝑘 for wide plates in compression

elastically restrained on the loaded edges
a<b
k  (a / b) 2 k

 The corresponding coefficient in the
"design" version of the wide plate
formula 2
t
( a )cr  KE  
a
 In ship structures the rotational restraint

is usually provided by flange-and-web
type of transverse stiffeners.
 In this case ζ is given approximately by
27a   2 Id 2 J 
  3 2 2  
tb  b 2.6 
d : depth of the web

I : second moment of area of the stiffener about
the midthickness of the web
J : Saint-Venant’s torsion constant for the
stiffener Buckling coefficient k for plates with
loaded edges simply supported and
34 longitudinal edges rotationally
OPen INteractive Structuralrestrained
Lab
12.3 Biaxial Compression
All Edges Simply Supported
 a is parallel to σax and b to σay. Aspect ratio α=a/b.
 Applying the energy method yields the following expression for the critical
combination:
 4t  b ax 2 a ay 2   4 at 2 m
2

W  C 

2
m  n 
   C 
mn    n 2
 
ay 

mn 2 ax
8 m m  a b  8b m m 
2 2
 ab4
2 m
2
n 2   4 ab 2 m
2
2
U D Cmn  2  2   D  C 
mn   n 

4 2
8 m m a b  8b m m 
 U=W  m  2   2 D  m 
2

2
  ax   n 2
 ay  =    n 2

    cr b t   
2

 If we denote the square plate critical stress and nondimensional form

2
t  m 2  ax    2

2
( ax )cr ,1  3.62 E     n 2 ay 1  m 
 =    n 
2
b 
  ( )
ax cr ,1 (  cr 4   
ax cr ,1 
) 

2
t
σay ( ax )cr ,1  3.62 E  
b
b α=0.5,m=1,n=2
a
b σx
a
α=1,m=1
σay
σx
b α=1,m=n=1 b α=3,m=3
a a
σay
b σx
a
b α=3,m=3
a α=0.5,m=1
When σax ≈σay, α=1

σay
m=1 m=2 2
t
m=3 ( ax   ay ) cr  (0.5  0.5)  3.62 E  
b α=3,m=1 b
a
2
t
 3.62 E  
b
Buckling stresses of biaxially loaded

simply supported36 plates OPen INteractive Structural Lab
Plate under biaxial load
a/b = 3 a/b = 5

DNV Rule for Classification of Ships Part 3 Chapter 1, Section 13
B400 Plate panel in bi-axial compression
 For plate panels subject to bi-axial compression the interaction between the
longitudinal and transverse buckling strength ratios is given by
n
 ax  ax ay   ay 
K   1
 x cx 
 x y cx cy   y cy 
Homework #2 Plot DNV bi-axial inte

raction curve and compare with the
previous interaction curve (Fig. 12.8)

All Edges Clamped
 For plates subjected to approximately equal compressive stresses

(σax ≈σay)the interaction formula is
2
t  3 
( ax   2 ay )cr  1.20 E    2  3 2  2 
 b   
2
t
 When α=1, ( ax   ay ) cr  9.6 E  
b
 For square plates(α=1), critical combinations are given for particular

values of σax / σay, including cases in which σay is tensile.
 When σax = σay

2 2
t t
( ax   ay ) cr  (5.61  5.61)  0.905 E    10.15 E  
b b

12.4 Other Types of In-plane Loads
Pure Shear
 In ship structures the plating is commonly subjected to large shear loads.
The shearing load can cause buckling since it gives rise to in-plane
compressive stress.
 For the case of pure shear, in-plane
compressive stress is equal to the shear stress
and acts at 45° to the shear axis.
N x  p  N y  0, N xy  t 2 t  2 w
 w
4
D xy
 In shear buckling, the coefficients are denoted
as ks and Ks.
2
 D
2
t
 cr  K s E  
 cr  ks
b 2t  
b
 For simply supported plates
ks =5.35+4(b / a) 2 Buckling of an infinitely long, simply
 For clamped plates supported plate
k s =8.98+5.6(b / a) 2
Pure Shear
 For simply supported plates
 2D ks =5.35+4(b / a) 2
 cr  ks
b 2t
2
2 2
t t
 cr  k s  
E  0.90 k s  
E
12(1  v 2 ) b b
B300 Plate panel in shear

(DNV Rule for Classification of Ships Part 3 Chapter 1, Section 13)
2 2
 t  s
 el  0.9kt E   ( N/mm ), kt  5.34  4 
2
 1000 s  l

Pure Shear
 ks and Ks are given for various types of boundary conditions. Because of the
symmetry of the pure shear loading , the choice of a and b is independent of
the load.
Buckling coefficient of flat plates Buckling coefficient of flat plates

in shear in shear
OPen (Design Structural
INteractive formula) Lab
Biaxial Compression and Shear
 For long plates ks is given approximately by:

– All edges simply supported:
1/2 1/2
   ay 1/2  ax     ay 1/2  ax 
ks =  2 1   2    2 1   6 
   e   e     e   e 
 
– All edges clamped:

1/2 1/2
 4   ay 
1/2
8  ax   4   ay 
1/2
 ax 
ks =   4       4  8 
 3   e  3  e   3   e   e 
 
 2D
where  e 
b 2t

In-plane Bending
 σb denotes the largest or edge
value of the applied stress.
 2D
( b )cr  kb
b 2t
 Some approximate formulas to calculate the values of kb

– simply supported edges:
for a/b 2/3 kb  15.87  1.87(b / a)2  8.6(a / b)2
for a/b  2/3 kb  23.9
– clamped edges:
for a / b 1 kb  41.8
– one unloaded edge clamped; the others simply supported
for a / b  1/ 2 kb  25
– unloaded edges clamped; loaded edges simply supported
for a / b  0.4 k  40
b

In-plane Bending
 The figure illustrates the case in which the bending is unsymmetric.

For simply supported edges the value of kb is given approximately by
kb  5 2 +4 (
a 2
 simply supported edges only)
b 3

Combined In-plane Loads: Interaction Formulas
 Uniaxial compression and in-plane bending

 (σa)cr : critical values of axial loading
 (σb)cr : critical values of and in-plane bend 1.75
 a   b 
   1
 ( a )cr 
  b cr 
( )
 Uniaxial load(compressive or tensile) and shear

 For convenience we adopt the symbol R to denote a critical load ratio.
In the present case the strength ratios are
a 
Rc  Rs 
( a )cr  cr
 The interaction formula is
Rc  Rs2  1  1
 1  0.6 
  Rc  Rs  1
2
 1
 1.6 
 In-plane bending and shear
b
Rb  Rb2  Rs2  1 ( >1/2)
( b )cr
 Biaxial compression, in-plane bending, and shear

 The two compression strength ratios are
 ax  ay
Rx  Ry 
( ax )cr ( ay )cr
 By performing a series of four-variable curve-fitting solutions,
0.625(1  0.6 /  ) Ry Rs2

 1  1
 Rb  1  Rx
4
(1  0.625Rx ) 1  2
 (1  Rx ) 
B500 Plate panel in bi-axial compression and shear
n
 ax ay   ay 
2
 ax  
K   1 q  1   a 
 x cx q  x y cx cy q   y cy q 

 a 

Interaction curves for biaxial compression, in-plane bending, and shear drawn for α=2

Homework #3 Plot DNV bi-axial interaction curve like

the right figure and compare with the following curve
for Rb=0
0.625(1  0.6 /  ) Ry Rs2

 1  1
 Rb  1  Rx
4
(1  0.625Rx ) 1  2
 (1  R x ) 
B500 Plate panel in bi-axial compression and shear

n
 ax  ax ay   ay  2
K   1  
 x cx q  x y cx cy q   y cy q 
 q  1   a 
 a 

12.6 Ultimate Strength of Plates
Plates Without Residual Stress
 Uniaxially loaded, simply supported square plate, with sides free to
pull in. some typical initial distortion in the form of a half wave in
each direction.
 Plate slenderness   b  Y
t E
 The relationship between the applied load (σa) and the axial
shortening
Plate strength without welding (σr=0)

 Slender plate (β>2.4)
 Buckling stress is well below yield stress and below the curve of collapse
stress.
 After buckling (σa) a greater proportion of the load is taken by the region of
plating near the sides → Non-uniform compressive stress distribution
 Deflected shape of the buckled portion → overall stiffness of the plate
(dσa/dεa) is reduced.
 The center region becomes more pronounced and the maximum stress at
the sides increases. When the maximum stress = yield stress → collapse.
Ultimate strength
Large margin between

buckling and collapse
Plate strength without welding (σr51

=0) Post-buckling
OPen INteractivestress distribution
Structural Lab
 Plates of intermediate slenderness (1<β<2.4)
 Buckling stress ≈ yield stress
 For a rigorous analysis, elasto-plastic large deflection theory to be used.
 As applied stress increases → magnification of the initial distortion → loss
of stiffness → some local yield → stress redistribution → yielding of the
sides → sudden collapse.
 Pitched roof : allows large axial shortening with minimum strain
energy.
Typical post buckling behavior

 Sturdy plates (1>β)
 The initial distortion is smaller and the magnification is less because the
elastic buckling stress is very large.
 Plates can carry a load equal to the full “squash load” σa,u= σY.
 After the peak load, the load carrying capacity remains
approximately constant up to very large strains.
Plate strength without welding (σr=0)

Plates With Residual Stress
 Departure from linearity occur at the stress which is less σa Idealized residual stress
distribution
less than for a stress-free plate.
 Sturdy plate (1>β) : no load shedding, but large
reduction in stiffness → regarded collapse.
r 2
 Intermediately slender and slender plate (1<β) : the =
 Y b  2
loss of ultimate strength ≈ σr t
Middle part
Edge part
σa + σr = σY
σa - σY = σY
Plate strength with welding (σr=0.1σY)

Effects of Other Parameters
 Restraint at Sides
 Clamping the sides of a plate increase the elastic buckling stress by
75%, however, the increase in buckling stress even in slender plate ≈
10% at most.
 Stiffeners surrounding the panel is not clamped edge → this restraint
can be ignored.
 Initial Deformation
 The effect of initial deformation removes sharp knuckle in curve of σa
and εa. The increasing lateral deflection causes a progressive reduction
in the in-plane stiffness of the plate.
 However, the ultimate strength is slightly decreased.
 Shear stress
 In –plane shear stress tends to lower the resistance to longitudinal
compression.
2
 Reduced yield stress rτσY     eq   x2   x y   y2  3 xy2   Y
r  1  3   
 Y

Ultimate Strength of Uniaxial Loaded Plates
 Plating of uniaxially loaded, longitudinally stiffened,
initial deformation (δp < 0.2βt), residual stress (σr ≈ 0.1σY)
side constrained to remain straight but free to pull in
 For sturdy plate, first loss of stiffness is taken as collapse.
 For plates of greater slenderness : loss of stiffness is gradual.
 Secant modulus ratio Es  10.4  2.75
T  
 0.25 2      2 ,   1  2
2
E    
Design curves of ultimate strength and secant modulus

Ultimate Strength of Uniaxial Loaded Plates
 Faulkner’s formula for the ultimate strength of unwedded plates : good
agreement with extensive experimental data.  a ,u 2 1
 
Y  2
 The effect of residual stress → strength reduction factor Rr
 E Ets 2   1
Rr  1  r ts  (1    2.5)
Y E E 
for   2.5 Ets  E and for   1.0 Ets  0
 Restrained : the sides remain

straight and do not pull in.
 Unrestrained : both types of
transverse deformations can
occur.
 Stress relived (σr=0), average
welding (σr≤0.1σr), heavily
welded (σr≤0.33σr)
Curves for ultimate strength of plates


Topics in Ship Structural Design: Lecture 5 Buckling and Ultimate Strength of Plates

Uploaded by

Copyright:

Available Formats

Topics in Ship Structural Design: Lecture 5 Buckling and Ultimate Strength of Plates

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Topics in Ship Structural Design: Lecture 5 Buckling and Ultimate Strength of Plates

Uploaded by

Copyright:

Available Formats

Topics in Ship Structural Design

(Hull Buckling and Ultimate Strength)

Lecture 5 Buckling and

OPen INteractive Structural Lab

Slide 2 OPen INteractive Structural Lab

Slide 3 OPen INteractive Structural Lab

OPen INteractive Structural Lab

 Unstiffened Plate (Plating between stiffeners)

Johnson-Ostenfeld plasticity correction formula

OPen INteractive Structural Lab

Simple support along the loaded edges

With no support along the unloaded edges

Buckling of wide column

 Large-Deflection Plate Theory

1 2w 2w 2w 

OPen INteractive Structural Lab

OPen INteractive Structural Lab

Load applied Elongation in V 1

E  2w 2w   2w 

OPen INteractive Structural Lab

For unit-width strip in Section 9.2 dx

OPen INteractive Structural Lab

 Because of W=U, and hence, 2 m

Q: Which critical stress will be higher?, which stiffener

OPen INteractive Structural Lab

 For long simply supported plates it is  2D

OPen INteractive Structural Lab

a/b=1, m=1 a/b=2, m=2

16 OPen INteractive Structural Lab

 For a general "wide plate“, in terms of a a

 For design purposes it may be written as:

17 OPen INteractive Structural Lab

 Transverse stiffeners: a<<b, a=s, b=B,

 Longitudinally stiffened plating have the great advantage over

18 OPen INteractive Structural Lab

Slide 19 OPen INteractive Structural Lab

Hull forces: Shear force and moment

Aft ship out of water

Wave crest around midship

Vertical ”g” force

OPen INteractive Structural Lab

21 OPen INteractive Structural Lab

 The buckling capacity might not have been checked sufficiently

OPen INteractive Structural Lab

OPen INteractive Structural Lab

OPen INteractive Structural Lab

OPen INteractive Structural Lab

 Increase buckling strength

OPen INteractive Structural Lab

Q: Which edge is more effective to in-

Buckling coefficient k in the design formula

 2D Unloaded edge : clamped

Unloaded edge : free & clamped

30 OPen INteractive Structural Lab

 Lundquist and Stowell have

 The degree of rotational restraint

 Cy : rotational stiffness of the

Buckling coefficient 𝑘 for wide plates in compression

33 OPen INteractive Structural Lab