f.rit--fjlAl{.-.,.

"n#'rni.rtiUr \i.b tIJ \rtd')..riL,- !tt6"

pi,.rr.lr;r-

tul+"Jg\,)t;r
"

w ori..:li;Keicj.;-g

"i rlp

eluK:t[

ru-f

'

Contents
Title
Acknowledgment
Dedication Contents
Page No.

tl

Il
IV

Chapter One- Introduction

1-1
L-Z

General
Features Aims of The Study

t
6

L-3

chapter Two-Problem DeveloPment

2-l
2.L.L
2-7-2

Design Method of Concrete Tanks Using Charts and Tables

7 7 8 9 10
11

Grillage Method Finite Element Method Comparisons

2-t-3 2-t-4
2-L-5

Developments
Basic Relations of Plate Analysis

2-2
2-2-L 2-2-2

L2 13

Thin and Thick Plates

Stress-Strain Relations

t4
15

2-2-3 2-2-4
2-2-5

Moment Curvature Relations

Equilibrium Equation
Plate Equation

16 18

2-3 z-3-L
2-?-2

Finite Element Approach

Assembly of Structure Stiffness Matrix Nodal Forces

20
30

31

2-t-3

Boundary Conditions

31

Chapter Three-Staad Pro Program

3-1

Introduction
Plate and Shell Element

?2
33 33

3-2
3-3

Geometry Modeling Considerations Element Load Specification

Element Local Coordinate Svstem

3-4
3-5 3-5 ?-7

34 39 39 45

Outout of Element Forces

Element Numbering

Chapter Four-Application

4-t
4-2

General
Fixed support

49 50 60

4-t

Pinned support

Chapter Five-Conclusions and Recommendations

5-1 Conclusions 68

5-2

Recommendations

69

Introduction
These tanks very important public stmctures' use' for drinking and washing' domestic water of storage for used may be swimming pools, sewage sedimentation ' "etc'

1-1 General

Water storage tank

axe

classified

known Water retaining shuctures commonly follows:according to it is position as

as utanks"

are

l23-

Tanks resting on ground'

Tanks constructed underground' Elevated tanks.

rectangular, All the above three tyPes of tanks maY be of circular, These tanks may be with or hexagonal or anY other shaPe in plane.

without a roof covering.

pressure from water in The side wall of ta.nks are subjected to water may also subjected to earth case of tanks restmg on ground while they

the

pressure (saturated or unsatuated) from outside

tanks.

in

case

of underground

of full of water While designing underground tanks' both conditions full' water pressure of or empty have to be considered' In case of tanks there is no soil outside of the side walls has to be considered assuming soil pressure is to be wall. However, in case when tank is empty' up lift from outside and also the stability of tanks against
considered maintained' pressure due to gtound water has to be

(r

Chapter One

lntroduction

or perforated masonry
together or
a

consist of solid Elevated tanks are supponed on staging which may walls or reinforced concrete columns braced

thin hollow shaft.

load and self The tanks portion is designed for water pressure' live resist wind forces weight of different parts, while the staging is design to
is generally provided and earthquake forces. In addition, slab in such cases or seepage is a as raft or piles depending upon soil conditions' Leakage it' impervious common problem in water retaining structures, to minimize important three concrete must be used. Also we have to consider a very factors in designing a reinforced concrete tanks, which are they
1- Sfength
:

2- Water tightness 3- Overall stabilitY

It is comrnon practice to use reinforced or pre-stressed concrete

structure for the retention exclusion or storage of water and other aqueous

liquids. Concrete is generally the most economical material of

construction (the word concrete has come to symbolize strength, stability

and the image of being set as stone) and when correctly designed and constructed

will provide long life

and low maintenance costs so not

surprising the concrete is the most common solution for storage tanks in

the water and waste water industry. Beyond the inherent durability of concrete and the variety of post-tension and corrosion protection
techniques are providing even longer- lasting and economical solution for
storage tarks.

Chapter One

Introduction

for the three general t)?es The following figures are simple dnwing
of tanks mentioned above '

Fig. (1-1) Ground Tank

Introduction

Fig. (1-3) Elevated Tank

On the other hand, there is the Spherical shaped storage tanks (Fig.1-4) which are generally used for storing products at pressures
above 35 kPa (ga).

A spheroidal tank (Fig.l-5)

that

is essentially spherical in shape except

it is somewhat

flattened. Hemispheroidal tanks have cylindrical

are

shells

with curved roofs and bottoms. Noded spheroidal tanks

generally used in the larger sizes and have intemal ties and supports to keep shell stresses low These tanks are generally used for storing
oroducts above 35 kPa.

Introduction

Fig.(1 -4) Spherical tank

Fig.( 1 -5) Spheroidal tank

Also Underground storage is most advantageous when

volumes are

large

to be

stored. Underground storage

is

especially

advantageous for high vapor pressure products. Types of underground storage are:

1-

Caverns constructed in salt by solution rnining or

conventional rnining.

23-

Cavems constructed

in

nonporous

rock

by

conventional rnining.

Caverns developed

by conversion of

depleted

coal, limestone, or salt mines to storage e.

l-2

Features

l-Rectangular shape makes best use

of

available space

2- May be installed at or below ground level or on towers 3-Clear interior allows easy inspection, draining and painting

-IJI-

Introduction

ground

behavior of The aim of this project is to study the structure details of the water storage tanks The project discusses the

1-3

Aims ofThe StudY

example to design ground water tank. Also we have presets a simple By using STAAD'Pro small a rectangular water tank with its details'

(frnite element method) program we aim to design a rectangular tank by

and study the results obtained'

Problem Develo

Problem DeveloPment
2-1 Design Methods of Concrete Tanks

of rectangulax water many methods are available to deal with designing they are simple or how tanks. But these methods are different with how much methods are gtven accurate the results which are ordained Some of these
below. 2-1-1 Using Charts and Tables

Tanks have traditionally been designed by reference to published

tables derived from elastic thin plate theory'

These tables cover isolated rectangular panels with various proportions, and edge conditions and Ioading as shown in Fig.(2-l).

Interpolation
the assumed

is

required and the values cannot rePresent the

real

interaction between adjacent walls and the base. The base rarely Provides

fixity and it is not easy to calculate these affects accurately

- (..ill:|r.ir(9bt) - (crr.{rrd.nlx?&}

Fig.(2- 1) Rectangular wall of a tank with fixed edges

Engineers now have access to powerful computers and suitable

software which should enable them to design tanks more accurately as complete structures and show worthwhile savings' This has not
become a routine method however because the time spent generating the structure and

loadings and dealing with the output can be such that

Chapter Two many tanks can still

be

designed economically and quicker by hand'

It is pointless to aim for tle

extreme accuracy

of a "Perfect

of factors such as Analysis" for a concrete structure when the variability by as ground conditions and panel thickness can affect the results a practical 20Yo in some cases. The aim, therefore, is to develop
mrrch as

give results to between and quick theoretical analysis procedure that can
5Yo and 10%o of

a"Perfect Analysis"

The following methods can meet these criteria and show that a
balance can be struck between accuracy and practicality'
2-1-2 Grillage Method

This method subdivides each wall and slab panel of the tank into a grillage of rectangular beams which are connected at the panel junctions Fig.Q-z)' The beams are given properties relating to their orientation, spacing and the panel thickness' The support conditions are applied at the base panel nodes as springs which can be specified to suit

piles or the stiffiress of the ground. The loads are applied onto
beams.

the

Fig.(2-2) Grillage modeling of rectangular tark The above model produces acceptably smooth moment diagrams
and the results have an error

of less than

4Yo. They can be

plotted

as

{8

Chapter Two

Problem DeveloPment

showninthecut-awayviewFig.(2-3)oragainstsectionsthroughthe
structrrre.Thevaluesmustbedividedbythebeamwidthtoproducevalues
oer meter width.

of rectangular Fig.(2-3) Moment diagram for grillage beams concrete tank

time The grillage pattern can be made coarser to reduce the input model size but at the expense of some loss of accuracy and

and

smoothness

of the results plot This can be more practical for

large

multiple tank structures.

2-1-3 Finite Element Method

This

method subdivides each panel

into a mesh

of

small

elements. The element thickness

is specified and the supports

or springs

pressure can be added at nodes in the same way as for the grillage' The on the loads on the panels can be applied by specifring the intensity panel as a

whole Finite Element Analysis (FEA) programs can produce a

colored results value contour

plot

as shown in Fig (2-4)'

t -J

Chapter Two

Problem Development

Fig (2-4) Contour plot of rectangular tank

The results from FEA can improve

meter width.

on the

accuracy and
with values per

smoothness of the grillage method and can be presented

In some FEA progftlms the results plots are based on the average or centre of element value so it is important in these cases to use small elements at panel joints. The coarse mesh option therefore is not
always appropriate with FEA but the speed loading input ouiweighs this. The

of mesh

generation and high

FEA method can show very

local force values

at

point loads such as at pile supports so these need

to

be modeled carefully to reflect their real width and load spread.

2-l-4 Comparisons

The grillage method uses general purpose 3-D space frame

analysis software which is used regularly by engineers

in

many design

offrces, whereas the FEA programs are more specialized and are not used
as

generally.

So this

can limit its use to the larger

organizations. The

preparation of the grillage model is not as fast as FEA mesh generation because the beams must be defined for each beam spacing and panel thickness condition.

The

application of uniform loading

is

equally fast

by

both

methods but hydrostatic loading is rnuch faster by FEA by virtue of its

full

panel loading facility.

Pentium computer can now analyze the above

,rJ_--

Chapter Two

Problem DeveloPrnent

a reat rssue' so run times are no longer minutes 5 than less in FEA model is more familiar to structural The output from the grillage model to show the results need to be converted manually

engineers but

it

does

ofvigilance by the engineer' per meter width. This requires a degree

2-1-5 DeveloPments

further so that the The grillage method needs to be automated beam spacing and let the designer can speci! the panel width and output the results in a per computer calculate the beam properties and
meter width format.

to allow The loading input method also needs to be enhanced

globally varying loads to be applied to the parel
as a

whole'

FEA programs like Staad Pro which axe to be used for tank at the element analysis should be able to access and plot the results
boundaries and plot moment diagrams'

Both analysis methods

would

benefit from a library

tank models which could be modified or benefit multiplied by simple input from the engineer' The most valuable detailing however, would be the automation of the reinforcement
to process. The results could easily be fed through a post processor speciff calculate the requirements to specified code' The engineer could
could the main parameters and preferred spacing and the computer review The produce reinforcement proposal diagrams

of

then

for

information could then be passed to the detailing program to produce shown in drawings to a standard format similar to the wall elevation
Fig.(2-s).

("

Fig'(2-5) Detail drawing

main tank analysis' It is necessary for holes and items not covered by the process for many tanks quite possible therefore, to automate the design

The details would then be checked and edited where

and bring

cost benefits to the water industry'

2-2 Basic Relations of Plate Analysis Plates are flat structures in which the thickness (t) is small compared

to the other two

dimensions, viz., the length

a and breadth b' An

infinitesimal element under a system

of normal and shearing group of

Fig'(2-6)' The thickness (t)

stresses isolated from a loaded plate is shown in

ofthe plate is along z-axis

(rz

Fig.(2-6)

system ofnormal and shearing group

of

shesses isolated from a loaded plate

2-2-1 Thin and Thick Plates

Geometrically, a plate is said to be thin

if the thickness ratio t' (=t/b'

where b is the least dimension of the plate, b<a) remains normally less than
1/20. Plates having t'value greater than 1/20 are normally called thick '

In thin plate theory, the stresses oz ' rxz (rzx) and qz (=zy) which are directed towards the z-axis across the thickness, are considered to be of negligible magnitude compared to the other stresses viz', ox, 6y and rxy (=tyx), and are disregarded. For all practical purposes, the plane of the thin
plates remains as unstressed or neutral and serve as the reference plane'

In thick plate theory, the stress components az, rxz arrd ryz are taken

into consideration since they are of comparable magnitudes and cannot be

neglected. The various in-plane stresses ox, oy and

rxy(1x)

do not remain

proportional to the distance z measured fiom the neutral plane through the variation of the stresses rxz as well as ryz across the thickness can still be
assumed

to be parabolic (as assurned in the

case

of

straight beams of

{ta

Problem DeveloPment
of thick plates are to be carried rectangular section). ln essence, the analysis plates can handled with a 2-D out using a 3-D analysis whereas that of thin
analysis.

2-2-2 Stress-Strain Relations

terms of stresses In 2-D problems, Hooke's law expressing strains in of the D matrix can be can be expressed as : E = D o in which elements

written as :

Dr7 = D22

=i
(2.t)

D7Z=Dzl=-;
D44

=r(T')

=:
(2.2a) (2.2b)

That is

ex ey

=!1ox

-ttotl

=llot -ttoxl
=:-;-ltxll
2(

yxy

L+lt) ,

(2.2c)

of yxy' yields solving ox, oy in terms of ex and ey and rxy in terms

o, = #
oy =

lex

eYl

CLIex

ClZ

eY

(2.3a)
(2.3b) (2.3c)

jplrt * ex) = c!2 ex t c22 tY

"ry=t';1t*tf=c44YxY
Where the

Cij

coefiicients are the elements

of the generalized

modulus of elasticity mahix C ( C =

D-r ) so that

#ra

Chaoter Two

Problem DeveloPment

Cll=C22= El(

l'vz)
(2.4)
G_r)2)

.,.-"r.--l!wL. _ vLL _

C4a=Ellz(r+v)l
The strain-curvature relations are
:

EX=-zTF=z^x
EV=-Z-:-=Zn! oy02a

Q.5a)
(2.sb)

vxv=-22?=-22K," oxoy
curvatures can be given as
:

(2.sc)

ofthe On introduction (2.5) in (2.3), the stresses expressed in terms

62a, Ez.A2a ox=-l.i167+v6;rl

(2.6a)

Ez -d2@ 02u, oy = -7-v2l-;F + v6;;l .^t E(L-!)z (l_vz)

r a2@

(2.6b)

LAxAy

(2.6c)

2-2-3 Moment Curvature Relations

from The magnitude of the various bending forces can be evaluated
the stresses referring to Fig.1.2 (c) as : (1'4a),(1'4b)and(1'4c)'

Introducing the expressions of 6x,6y and

ay

ftom (2'6)
:

'

the

moment -curvature relations can be given as

Problem DeveloPment

ux = lttlrox
=

z dz

1-v2'Ax

"*(;+,te;fi,1,,,

a"
Q'la)

-rl# *

DlKx + Y

KYf
+'
uleS

"?r)=
z

uy =

f!,j,oy

a, = -

fi ffi
+v

f!,',"'

a"

-rlrtu*'#l=

DIKY

Kxir
=##:,tS
I!!rt,

Q'lb)

with rxr-rln<,

Mxy = 1v1r, = !'/rz,r-txv

z dz

"t

a"

ft -'>ltz)

= DG

- v)lKxYl

(2.7c)

where D =

#6=

Plate rigiditY

Q.7d)

2-2-4 EquilibriumEquation
moments act per unit length of Noting that the transverse shear and the plate equations for the loaded flat the element ( Fig. 2-7 ), the Equilibrium
elements are
:

l.
le,

Fz (sum for forces along z) = 0 gives:

fiom x+ and

x-

faces

*Y

o4o,
Y-

lQxldY

or ff)a'ar

from Y+ rmd

faces

lot +'g

av]ax

-lQYldx or

lffia*at

Chapter Two

Problem Development

Ay++

l--

dv

Fig.(2-7) transvene shear and the moments act per unit length
the element

of

From extemal loading

lqltlxdy
Adding these together with dx dy as common

a3'a93!as=g
Ax
AY

(2.8a)

2.

ZMy (sum of moment about AB ) = 0 given: From r+ond x- faces

luu

uff a,lav - lMxYldY

Yfaces

"' fffla*
+ff
av

at

The contribution fiom Qx and Mx are absenl'

From y+cnd

-Iuv +Yaylax av "t

faMr'la16+

+ lMy\dx +lOt

+ qv\ax!

Qtlaxay

The contribution from the extemal loading q and Mxy(=Mxy) are

Problem Development
absent and the product
neglected.

of smaller quantities larger than

second order is

Adding these together with dx dy as common, 1O,

t--

_ayv *at!-'v = ox dy

o1
.t

(2.8b)

3.

lMx(sttmof

moment qbout CD ) = 0 given:

Fromx+and.

x- laces -lur +'ffarlay + lMxldy +lQ, + Q, +ff axlati

-^l - f AMx + or [- ar Qx]axdy
The contribution ofthe extemal loading q and Mxy are absent are the

product of smaller quantities larger than second order is neglected.

Adding these together with dx dy as common,

0 ox-Y+Y= .lx oy
(3
.1

(2.8c)

2 a-c) are the three required equilibrium equations.

2-2-5 Plate Equation

The three equilibrium equations can be combined in to a single

equation on elimination Qx and Qy in (2.8a) using (2.8c) and (2.8b) as follows:
From (2.8c)

^ = "'.Ux ctx-'=oy

AMx

Myx

AQx A2M, 02uyx 0x Axz AIAY AMy Mty ^ = -'-UV -'--:-"dYdx AQy A2My Ay 0y2
02MxY axAY

(2.9a)

(2.eb)

Using (2.9 a,b) in (2.8a) and noting that

a2Mx , dzMy na2Mxy , .-t":---z-:--:--q oxoY oY' dx' ^

Mxy:M)t(

=0

(2.10)

{18-

This is

the single equilibrium equation

of

rectangular plates of
oan be

q=q(x,y) which uniform thickness under continuous loading

relation (2.7a-c)

exoressedintermsofthedisplacementt,-lusingthemomentcuwature

Now ,:lz=

-o(;+,

*iY,r")

t+---nffi+"ff";
-, !##
= - D t(2 - tut iuY,, u, )
:

(2'10) Adding these together and substituting in

-olffi+2ffi+aff+t=o
or
vaa = q/D
as

(2.11a)

(2.1Ib)
Sophie Germaine's equation( 1815)'

This equation is knoun

In non-dimensional form using

,=i,v =i
Where, a and
respectivelY, so that

(2.1

ic)

are the dimensions

of the plate along x and y

Aaa I Aau ax1 a2 Ait

67;7=

Aaa

L
-'zuz

Aaa

*zu*

(2.1id)

aaa L aaa ay1 b+ at4

We get the plate equation (2'11a) is

n, .6au oaa "#+zc'- a*,N;+"' dr=-,

a4

Where c2=constant=4z /b2

(2.1 I e)

trs

Problem Deve In (2.1lb) the operator Va is known

as

biharmonic operator'

aa 0a -'> tai ^a r#^L,rn.* = v'z(v') vr= #+

(2.12a)

Where V2= laPlace olerator

ax2 '

a2

a2 AYz

Q.12b)

2-3 Finite Element Approach

plate is divided into a series of small In the finite element method' the points only and then joined at their nodal elements. These elements are (Ref' 5)' is established at these points continuity, together with equilibrium'

co-ordinate and

for the rectangular element' the To derive the stiffiress characteristics in Fig'(2-8'a)' are used The node numbering systems shown

elementhastkeedegreesoffreedomateachnode,namelyt}vorotationsand

thetransversedeflection,i.e.0x,0yandw.thepositivedirectionsofthese
corkscrew rule' The element rotation are defined according to the right-hand fteedom as shown in Fig'(2-8 b) the then has a total of twelve degrees of of two moments Tx and Ty and a corresponding moments and forces consist
(2-8 c)' force Fz at each node, see Fig

For example, for node (1) we can

wite

the displacements

asi

{d1} =

lsJt]

[,0

Chapter Two

Problem DeveloPment

(A

(B

JY1

Tx

Fig.(2-8) continued

aart,

(D
element:Fig.(2-8) rectangular plate bending

A) B) C) D)

Co-ordinate system' Nodal displacement' Nodal forces' Intemal moments'

and forces as :and the corresponding moments

(rr1)

trt] = lrYtl \Tzl)

and force vectors for the element So that the complete displacement
can be written as :-

(0xr\

l?il lex,
{6'\
A"d {6"} =

= lwz \ 0x.

ltt;
lrr.
lw" le*,
loY'

tliJ

Problern DeveloPment

tTxl\

lrvt I lrzt I
lTxZ
I

lrvzl o"61r"l=Jr:l= lrzz\ lrx3 |

rF1)
|

?v3 |

lFz3l
I

lrx4l
rv4 | (2.13 )

{Fe} = [Ke]{de}
Where Ke ts

the l2xl2 stiffttess matrix of the element' Since

a suitable function is chosen which element has twelve degrees of freedom'

has twelve undetermined
1p

=1]io.2

+o(3 y +o(4 x2
ys

+e,

xy +xu y2 +x7 x3 +
(2'14)

dB x2y

+(s x yz +d.ro

+x[

x3y

+x,t xy3

following expressions for the The displacement function gives the

rotations :-

Aw xy+3x',o !2 0x=-''=-(o(3+o(s xt2d.ay]id-s X2+2d's tc(rr 13 r 3 xn x!2)

=a{ =xr+ 3 4r xzY +efl Y3)

0y

2o(4

+o(s y + 3 67 x2 + 2 eB

xy +ee yz +

Equation (2.14) can be written in

matrix from as :-

:lt ::l
(0x]

t'JJ=

loo -1 o-r-2l0 Ior o 2r v o 3rz Itx y x' xY Y"'

o 3tzy

Y'l

Q.1s)

-..l

,3

F-

Chapter

Two

Problem DeveloPment

Which can be summarized in generalmas:-

{d(r,y)} =

lf@,fl1{;a}

(2.16)

Then, the substitution of the nodal co-ordinate values into equation

(2.16) leads to the formation ofthe [A] matrix as :-

[de] =

[,4]{a}
for

(2.r7)
the element has a total of

This [A] matrix is a l2xl2 matrix

twelve degrees of freedom:-

IAI =

0 0 0 0 0 0 0 00 010 0 00 0 0 0 0 0 0 00 10 0 0 0 0 0 0 0 0 0 00-1 0 0 -2b 0 0 0 -3bz 0 o b3 01 o o D o o o b'z o 0 0 b3 0 0 10 ,0 0 b,0 0 00-1 0-a o 0 -a2 0 0 0 01 O 2a 0 0 3a'z 0 0 0 0 o o la o a, o o a2 o o o 0 0 ..1 0 -a -2b o -a2 -zab -3b'z -a3 -3ob'z or O 2a b 0 3a2 2ab bz 0 3a2b b3
0 0 -1
I a b az db bz a3 42b ab2 b1 atb

(2.18)

ab!

Nodei x=0 ,50

x=a
,

Node2 x=0 , 5b

Node3 x=a ,

5o

Node4

5b

From equation (2.17)

{d} = [,4]-1{de}
Substituting in to equation (2.16):{6 (x,

y) } = [f (x, y)llA]-

{6

(2.1e)
as, from

Then, the state

theory ofplates:-

of strain in the element can be represented

,o)

Chaptr Two

Problem DeveloPment

{e(r,y)}=l-a'*1ay'l
lz02w /axay)

( -A2w /0x2

(2.20)

And substituting for w from equation (2' 14) gives

{e(x,v)}=

(-azw/ax?\ ( | -6z116rz l=]

lza'*faiay)

-(2q.4+6C/'? x+2dB Y+6xr xv) -(2C/.6+2xe x+6rrc Y+6c(l' xv) (z1xr+z ea x + 2 ds v +3 (/-1t x2 + 3 xtz

,'r]

0 o -6xY o] [o0o-2 o o -6x -2Y o o -21t -6Y o -6xYl =fooo o o-z [ooo o z o o 4x 4v o 6xz 6v2 ]

(2.2r)

This above equation can be summarized as :-

{e(x,

y)} = [c]{a}

(2.22)

But from equation (2.17):-

[a] = [A]-1{de}
Then substituting in to equating (2.22):-

te(x,

y)j = [c] [e]-1{d'e}

Let tBl =
.'.

[c]Pl-l
Q.23)

{e(r, Y)} = tBlid"}

Next step, we must relate intemal stresses {o(x,y)} to strains {s(x'y)} and to nodal displacement [d"]. in a plate flexure solution the intemal

stresses are :-

( M.)

{o(r,Y)} = 1M, I

(2.24)

lM")

per unit length Where Mx and My are the intemal bending moments up within the and Mxy is the intemal twisting moment per unit length set
element, see Fig. (2-8 d).

From plate bending theory, the stress-strain relationships are:-

:2-,\ Dl;i) Mx = -lDx#+ ^2-,

Ftt -

,"---(oru2+otfu. ) \"t Ayz ' -- Ax2)

Mx! =
2D

-:-"-

"

y directions Where Dx and Dy are the flexural rigidities in the x and ratio O?e of respectively, Dl is a coupling rigidity representing a Poisson's
effect and Dxy is the torsional rigidity'

But for an isotroPic Platei

Dx=Dy=D=r=l-ur_,1

D7=vD
And

Dxr

1(1

v)D

form as:Then we can write equation (2 24)and(2'25) in a matrix

1,u*

Problem Deve

{o(x,y)}

=[Y',;l=lBI
\MxYJ
t0

r;,.lJl-lfl

(2.26)

to(x,y)j = [o]{e(x,y)}

(2.27)

This defrning the [D] matrix .substituting for {e(x'y)} from general
equation (2.23) gives the required relationship between element stresses and nodal displacement as

:-

{o(x'y)} =

[D]

[B][de]
statically

The intemal

shesses

{d(t,y)} are now replaced by

equivalent nodal forces {Fe}' the principle

of virtual work is used, where

during any arbitrary virtual displacements imposed on the element, the total
extemal work done by the nodal loads must equal to the total intemal work

done by the stresses selecting

a set of virtual nodal displacements:

{d.e} =

U}

(the identity column matrix).

The extemal work done by the nodal loads is (14l"'g),

wext = {6"e}r {Fe} =

{Ixre} = {Fe}
y)]

If

the imposed virtual

displacements causes strains [e-(x,

(curvatures) at a point within the element where the actual stresses are

{o(x,y)} (bending moments per unit length ), then the total intemal work
(tr4l;,"r) done

will be

ba
w,-,
Substituting,

=|
00

| {e'" (x,y)}r [o(x, y)J dx dy.

Chapter

Two

Problem Development

{e-(x,y)} = [B] {d.e} = tBltl} = tBl

And io(r, y)] = [D][B] {de}
Then

w,,,

: Il

fft9l'[D][B]

dxdy. {6e}

equation

W"rt:Wint
Then,

tF"j = (: I:lBlr

[D][B] dx dy

)6e

(2.28)

And the element stif&ress matrix [Kel is defined as :-

tK"l =

I:tBl',[D][B] dxdy

(2.2e)

Thus the final value of the slifftress matrix [Ke]Obtained from this
above equation

is as given in the following equation (2.30) for

general

orthotroDic case :5,{

-JA -SD

56

-st
-JD
JC
0

scsE0 JE Sf Srt
5/ 5./
tB

0 -sn sli .rr s./ 0

S,' JM JO

sa
JS
0

so sP o -55 0 sr

trl

=;;
0 JO

s./ sM sD 0sosP
SRJSO

-ss sx

..rQ

0 -.JQ .t/v ,l? su 0 sx .s.t 0 -.tu 5v -so -.ts sx -Jn

-sQ

sa sD .tP -sQ J,f/ 0 -s0 | 0 sr sc .tt -5u o sR -ss Sf SF JQ SU .t/ -.to s5 sx | 0 sQ J,{ JA -sD J6 0 -sril Jl .ru sa 5c -rE 0 5r -l,/| -.t, sr -sD -sd sr .rr -.r, s,r, ,, ,, -ru ,, 0-J0s60
I I

sx -sa .ru

,01 -sul
sv
I

(2.10)

Jt -s,/

-5.' -58 .tC

JM .'D _S'

I I

-SE

JF

Where:

=;

Cha tr Two

Problm Devlopmnt

SA SB

= 20 az Dy + 8b2Dxy = 15 ab Dt

sc =20b2 Dx*8a2 Dxy

.9D

=30

apDy+15b\+

6bDxY
6

SE

= 30bp-rDx +15 aDLt

aDxy

sF = 60 p-2Dx + 60p2Dy + 30 D]-+ 84 Dxy

SG

= l0

az Dy

2b2

Dxy

SH =

-30 apDy-6bDxy

SI=I0bzDx-8a2Dxy
sl =
SM 75 b p-Lox

- L5 aD1- 6 a Dxy
60p2 Dy

=30p-2 Dx -

30 D1- 84 Dxy

sN = 10 a2 Dy
SO

-qbz

Dxy
6 b DxY

= -15 p o Dy + 15 b DL *

SP=5azDy+2b2Dxy

SQ=73apDy-6bDxy
sR=10b2Dx-2a2Dxy
SS=30bp-1 Dx + 6 aDxy

ST=5b2Dx+2a2Dxy
su = 15bp-a Dx-6aDxy
sX

= -60 p-2 Dx *

30 p2 Dy

30D1

84 Dxy

Chapter Two

sY

= -30 p-2 Dx -

30 p2 Dy

+ 30':-

84 Dxy

2-3-1 Assembly of Structure Stiffness Matrix :-

After establishing the stifftress matrices of all the elements in the structure, they are assembled together so that a relationship is obtained
between the extemal forces acting at the nodal points ofthe structure and the displacements of these points.

Denoting all the extemal nodal forces by :-

And all the nodal displacement as:-

"fi
With N; and

Ui

representing the th,ree components of extemal forces

.

and the three corresponding displacement at the*I the node respectively

thus:

tsl {u}={N}
Where S is the overall structure stimless matrix . The elements of matrix S are built up by adding the stiffness of elements adjacent to a node
,as

follows:
sq=2.=".r,...x1t

Chapter Two
2-3-2 Nodal Forces

Problem Development

In most ofthe problems with in this work, the plate is subjected to a

uniformly distributed load of value "q", whereas in the derivation of the

stiffiress matrix , concentrated extemal nodal loads are assumed . Thus, the

lumped load representation is used to replace the uniformly distributed load to equivalent concentrated load at the nodes. 2-3-3 Boundary Conditions
Three types of boundary condition involved in this work, they are: the

fixed edge in which, the deflection and slope normal to that edge are equal
zero, and the simply supported edge in which the deflection and the rotation about axis normal to the edge are equal to zero , and the free edge.

1J

Chapter Three

Staad Pro Program

Staad Pro Program 3.1 Introduction

Staad offers general purpose structural analysis and design along

with

extensive model generation and post-processing facilities.

All

these

features are integrated in one common Graphical User Interface (GUI).

This manual describes the Staad GUI in detail. The Staad GUI has a concept called Page Control. When it is switched on
(using the Page Control option from the Mode menu), a tabbed menu
appears along the left side

ofthe screen

as a guide to the process of creating

a structure. Every "Page" serves a specific purpose. For example, the

General I Load Page offers facilities to define different tlpes ofloads. There are two levels oftabs indicating pages and subpages. You should start from the top ofthe tabbed menu and gradually work down through the
pages and subpages to input geometry, supports, loads, and other

parameters.

In addition to the GUI, STAAD also offers an Input Command File

interface for speciling the Input, Analysis, and Output commands. This

file is a text file consisting of simple EnglishJike commands. You develop

or modiff the model using either the Staad GUI or the Input Command

File. When a model is created using the graphical tools, the input command

file is automatically generated.

'Thlr chapter

is adopted

from Staad Pro r|elp

,,lt . 3
:

,)J;t:f{^$r, ;Wt rA Sa,.rit;i

;

r,,..

(;r g,l,jJ .l.. q4

"g ole.; er^+;ar(-.- y b r,"ri 0 (LiV,.G\ J.0tt ik-,V

..WAY,p'lilt).et;t

*6i4

a,;antlrsLV y,g-

JI

J/t n+:ltL'.;

;i 6 L<-

)V.<-,Jtb];l-,t.JbJlyp;:tuan"rr#t,xc.rb.nb*a!\,*n
:

ot,(.lyt"L.tt---;!-

U4r4r;i.irr;u!r.,,.

w-^\l)b
.llrlp"uLu.l

;s--iL,"(.litlei

Staad Pro Program

3-2 Plate and Shell Element

ThePlate/Shellfiniteelementisbasedonthehybridelementformulation. Theelementcanbe3-noded(triangular)or4-noded(quadrilateral).Ifall is of a quadrilateral element do not lie on one plane' it

the four nodes The thickness ofthe advisable to model them as triangular elements' element may be different from one node to another' and shells may be modeled "surface structures" such as walls, slabs, Plates ofa finer mesh of using finite elements. For convenience in generation

Plate/Shellelementswithina|argexea,ameshgenerationfacilityis
available.

action only The The user may also use the element for plane stress
for this purpose' element plane stress command should be used

3-3 Geometry Modeting Considerations

be remembered The following geometry related modeling rules should
while using the Plate / Shell element'

l-

node "O" (center node - see The program automatically generates a fifth

Fig.3.1) at the element center' the input data' it is essential that 2- While assigning nodes to an element in

thenodesbespecifiedeitherclockwiseorcounterclockwise(Fig.3.2).For
numbered sequentially' better efftciency, similar elements should be

They should be on the 3- Element aspect ratio should not be excessive

order of 1 :1, and preferably less than 4:1 '

kom sradd Pro Help 'Thi5 chaprer D adopied

33-

Chapter Three

Staad Pro Program

between two elements should not be distorted Angles larger than 90 and never larger adjacent element sides should not be much

4- Individual

than 180 see Fig(3.4). 3-4 Element

Load SPeci{ication

Following load specifications are available:

directions' 1- Joint loads at element nodes in global

the element m 2- Concentrated loads at any user specified point within

global or local directions. global or local directions 3- Uniform pressure on element surface in

portion of element surface in 4- Partial unifbrm pressure on user specified

global or local directions
surface in local directions' 5- Linearly varying pressure on element or decrease oftemperature' 6- Temperature load due to uniform increase

between top and 7- Temperature load due to difference in temperature bottom surfaces of the element'

.ThBchaprrk adopred kof, Staad Pto

Hlp

14i

Staad Pro Program

lr
lr

gof.ct

.l lr
|.t,rrb.rh0

rl
.l
q

l* ttll 'l l,
Jl lr
6enQm!9d

Corder ll!.le)

od?

l|forrict

tl|lberh9

Fig.(3. 1)Fifth node generation

Fig.(3.2)Correct element numbering

ttt/ lll/

-'--'-1
6ood Rmnas

\_Z x...
//\
ldd
El?t'r

r/' ,'.1

,,.

I
I

A
Fig.(3.3) Good element q?es

tA
Fig.(3.4) Bad element tYPes
, t 1s-

element The staad plate finite element is based on hybrid finite is assumed' For formulations. A complete quadratic stress distribution
as follows Fig(3'5)' olane stress action, the assumed stress distribution is

.Thischapreriradoptedfiomst.adProH.lp

Staad Pro Pro

(l\

Fig.(3.5) Plane stresses Complete quadratic assumed stress distribution:

lt, \oxy /
tl

/ o,\
I

4rr
0

x y00 = t; 0 01r Io -/000

41

-x | -Zxy -y2 0 0 -xzl

o xz 2xy y2 o ol o y' o o xz zxyl

t;\

OrlI

:J
a"z,/

through dlz =constarts of stress polynomials'

The following quadratic stress distribution is assumed for plate bending

action Fig.(3.6):

L"
Fig.(3.6) Out-of- Plane action
1flit
chapter h adopted frcm staad P,o Help

[,u

Staad Pro

Complete quadratic assumed stress distribution:

x2 xYY2 O 0 0 tM,\ llxyO0OOO0 /ur\ lo 0 o 1 x y o 0 0 o 0 0 x2 xv v2 l,nt'" l=fo 0 0 0 0 o I x y -xy o 0 0 0 -xy o -t I q, I lo I o o o o o o 1 x v o o \Qyl [o o o o o 1 o 1 o -/ o o o x Y

al through
417 =constants of stress polynomials.

x2

0 0l 0 0l
y2l

;j\
:)

o zYl 2' o]

The distinguishing features of this finite element axe:

1- Displacement compatibility between the plane stress component of

one element and the plate bending component of an adjacent

element which is at an angle to the first (see Fig.3.7) is achieved by

the elements. This compatibility requirement is usually ignored in

most flat shell/ plate elements.

Fig.(3.7)Displacement compatibility
2- The out of plane rotational stiffiress from the plane stress portion ot each

element is usefully incorporated and not treated as a dummy as is usually

done in most commonly available commercial software'
'ThG chaptet ii.dopied fron staad Pro
He19

#'

Staad Pro P

3-

Despite the incorporation

of the rotational

stiffness mentioned

previously, the elements satisS the patch test absolutely.

4- These elements are available as triangles and quadrilaterals, with comer

nodes only,

with each node having six degrees offreedom.

elernents

5- These elements are the simplest forms of flat shell/plate

possible with comer nodes only and six degrees of freedom per node. Yet

solutions to sample problems converge rapidly to accurate answers even

with

a laree mesh size.

6- These elements may be comected to plane/space frame members with

full

displacement compatibility.

No additional restraints/releases

are

reouired.

7- Out of plane shear strain energy is incorporated in the formulation of

the plate bending component. As a result, the elements respond to Poisson

boundary conditions which are considered to be more accurate than the

customary Kirchhoff boundary conditions

8- The plate bending portion can handle thick and thin plates, thus extending the usefulness of the plate elements into a multiplicity of
problems. In addition, the thickness of the plate is taken into consideration

in calculating the out ofplane shear.

9- The plane stress triangle behaves almost on par with the well known linear stress triangle. The triangles of most similar flat shell elements incorporate the constant stress triangle which has very slow rates of
convergence. Thus the triangular shell element is very useful in problems

with double curvature where the quadrilateral element may not be suitable.
10- Stress retrieval at nodes and at any point within the element.
'rhl'
chaeter I' adoeted f@m staad Prc

lere f 38 I

t,

Chaoter Three

3-5 Element Local Coordinate System

The precise orientation oflocal coordinates is determined as follows:

l-

Designate the midpoints of the four or three element edges IJ' JK, KL,

LI by M, N, O, P respectively.
2- The vector pointing from P to N is defined to be the local x- axis. (In a triangle, this is always parallel to IJ).

3- The cross-product of vectors PN and MO (for a triangle' ON and MK)

defines the local z-axis, i.e., z = PN x MO.

4- The cross-product of vectors z and x defines the local y- axis, i.e., y = z

xx.
The sign convention of output force and moment resultants is illustrated in Fig.3.9.

Fig.(3.8)Element designation 3-6

Output of Element Forces

Element force outputs are available at the following locations:

A. Center point ofthe element.

I tlis chapte,
is adopted

from st.ad Pro Help

B. All comer nodes olthe element.

C. At any user specified point within the element.

Following are the items included in the element stress output.

SQX, SQY SX, SY, SXY
ND(, MY,
shear stresses (force/unit len./unit thk.)

membrane stresses (force/ unit len./ unit thk.)

N,Dry

bending moments per unit width (moment/unit len.)

SMAX, SMIN

principal stresses (force/unit area) maxim. Shear shesses (force/unit area)

orientation ofthe principal plane (degrees)

TMAX
ANGLE

VONT,VONB

Von Mises stress

Vtt=OIOZ@
Notes:
1.

All element stress output

ofthe element

is in the local coordinate system' The direction

and sense

stresses are explained in Fig.3.9

2. To obtain element stresses at a specified point within the element, the user must provide the coordinate system for the element. Note that the

origin of the local coordinate system coincides with the center node of the
element.

3.Principal stresses (SMAX

&

SMIN), the maximum shear

stress

(TMAX), the orientation of the principal plane (ANGLE), and the Von Mis Stress (VONT & VONB) are also printed for the top and bottom

'rhB(hapt., isadopred lrcm

staad eto

HPlp

f 40 I_-_-

Chapter

Three

Staad Pro Program

surfaces of the elements. The top and the bottom surfaces are determined on the basis ofthe direction ofthe local z-axis.

Fig.(J.9)sign conrenlion of Elemenl Forces

.ni,

rhapter ii adopted iiom Staad Pro

Help f

,r

ter Three

T.nsd

i3

6 ne bal

F d'' rtd 'f'e tfi' rocllrJdce s-,mddl.. o h.loelkstt rr@rvrae u'G roc'rvr'a trlf'#6;; ;"*,t* 'r'. lodl b rr'o rdis !iD.r'ddltr
M-E lh. B.rdn{r MorFr

lP

's
!s

the

rac'

thr

r'E

/t@al L-----*,local x

SaREss CausED BY Mx

c)

ov\ I

\(+zsudace)

stress Caused bY M
Fig(3.9) continued
'Thir ch.Dter ir :dopted from staad Pro tlelp

(D)

Staad Pro Program

mombrans Stresss Sx

e4l|!Y

(E)

tocal

To*sroN

(F)

/t'
**\**

./,

***fo
ln Dlane shear stresses
Fig(3.9) continued
-Ihis chapter h adopted lrom Staad

and s

G)

ProHelp

t-

Staad Pro P

Out-ot nlane sheat stresses

Fig(3.9) continued

(H)

Please note the following few restrictions in using the finite element

portion of staad:
1- Both frame members and finite elements car be used together in a staad

analysis. The element incidences command must directly follow the member incidences input. 2- The selfweight ofthe finite elements is converted to joint loads at the connected nodes and is not used as an element pressure load'
3- Element stresses are printed at the centroid and joints, but not along edge.

ary

4- In addition to the stresses shown in Fig'3.9, the Von Mises stresses at the top and bottom surface ofthe element are also printed'

.rhc

chapter ir adopted lrcm staad P'o

herp f
44

Chapter Three

Staad Pro Program

3-7 Element Numbering

During the generation of element stiffness matrix, the program verifies
whether the elernent is same as the previous one or not.

Ifit

is same,

iepetitive calculations are not performed. The sequence in which the

element stiftress matrix is generated is the same as the sequence in which
elements are input in element incidences.

Therefore, to save some computing time, similar elements should be

numbered sequentially. Fig.3.10 shows examples ofeflicient and noneffi cient element numbering.

However the user has to decide between adopting a numbering system

which reduces the computation time versus a numbering system which

increases the ease of defining the structure geometry.

EFFlcleht elehent nunbFlng

I
?

5 6
B

Ihef lclcnt eleneht nunberlng

Fig.(3. I 0) Element numbering

.ThG chapteris adopred

lroh sta.d

Pro Help

Chapter Three

Staad Pro Program

Example

This example illustrates the usage of commands necessary to utilize the built-in
generation facility to generate spring supports for a slab on grade. The slab is
subjected to vadous types of loading and analysis of the struature is performed. The numbers shown in the diagram below are the element numbe$.

STAAD SPACE SLAB ON GRADE

Every STAAD input file has to begin with the word STAAD. The word SPACE signifies that the structure is a space frame and the geometry is defined through

X, Y and Z axes. The remainder ofthe words form

proJect.

title to identify this

UNIT FEET KIP

The units for the data that follows are specified above.

JOINT COORDINATES

0.0 0.0 40.0

2 0.0 0.0 36.0 3 0.0 0.0 28.t67 4 0.0 0.0 20.333

'l1rb chapter
is adopted

ffom

st!.d

Pro Help

Chapter

Three

Staad Pro Program

5 0.0 0.0 12.5 6 0.0 0.0 6.5 7 0.0 0.0 0.0

REPEAT ALL 3 8,5

O.O O.O

REPEAT 3 8.0 0.0 0.0 REPEAT 5 6.0 0.0 0.0 REPEAT 3 8.0 0.0 0.0 REPEAT 3 8.5 0.0 0.0
For joints 1 through 7, the joint number followed by the X, Y and Z coordinates are specified above. The coordinates ofthese joints is used as a basis for

generating 21 more joints by incrementing the X coordinate of each ofthese 7

joints by 8.5 feet, 3 times. REPEAT commands are used to geoelate the
remainingjoints ofthe structure. The results ofthe generation may be visually verified using the STAAD graphical viewing facilities.
ELEMENT INCIDENCES

l l892TO6
REPEAT 16 6 7
The incidences of element number

is defined and that data is used as a basis

for genelating the 2nd through the 6th elemerlt, The incidence pattern of the Iirst 6 elements is then used to generate the incidences of 96
elements using the REPEAT command.

(:

16 x 6) more

UNIT INCH ELEMENT PROPERTIES

1

TO 102 TH 5.5

The thickness ofelements

I to

102 is specihed as 5.5 inches

following the

command ELEMENT PROPERTIES.

UNIT FEET
CONSTANTS
E 420000.

ALL

POISSON O.I2 ALL

llhis
chapte. is .dopted

froh Stard

Pro Help

Chapter Three

Staad Pro Program

The rnodulus of elasticity (E) and Poisson's Ratio are specified following the

command CONSTANTS.
SUPPORTS
1

TO 126 ELASTIC MAT DIRECTION Y SUB

IO.O

The above comrnand is used to instruct STAAD to generate supports with springs which are effective in the global Y direction. These springs are located

l0 kip/cu.ft. The program will determine the area under the influence of each joint and
at nodes 126. The subgrade modulus ofthe soil is specified as

I to

multiply the influence area by the sub grade modulus to arrive at the spring
stiffness for the "FY" degree of freedom at the joint. Additional information on

this feature may be found in the STAAD Technical Reference Manual. PRINT SUPP INFO
This command

will

enable us to obtain the details of the support springs which

were generated using the earlier commands.

LOAD I WEIGHT OF MAT & EARTH

ELEMENT LOAD

I TO 102 PR CY -1.55
The above data describe a static load case. A pressure load of 1.55 kip/sq.ft

acting in the negative global Y direction is applied on all the 102 elements.

LOAD 2 'COLUMN LOAD-DL+LL'

JOINT LOADS

| 2 Fv -2t7.
8 9 FY -109. 5 FY -308.7 6 FY -617.4

22 23 FY -410. 29 30 FY -205. 26 FY -542.7 27 FY -1085.4

.rhischapter

i5

adopted ftom staad P'o

rielp

TJ

Chapter Three
43 44 50

Staad Pro Program

5t

',1t

7218 79 Fv -307.5

47 54 82 FY -264.2 48 55 76 83 FY -528.3
92 93

Fv -205.0

99 100 FY -410.0
103 FY -487.0 104 FY -974.0

113 114 FY -109.0

r20 r2r FY

-217 .0

r24 Fv -273.3
125 FY -546.6

Load case 2 consists of sevelal joint loads acting in the negative global Y

direction.

LOADING COMBINATION lOI TOTAL LOAD

| 1.2 t.
A load combination case, identified with load case number 101, is specified
above.

It instructs STAAD to factor loads I and

2 by a value

of 1.0 and then

algebraically add the results.

PERFORM ANALYSIS The analysis is initiated using the above command.

LOAD LIST IOI

PRINT JOINT DISPLACEMENTS LIST 33 56 PRINT ELEMENT STRESSES LIST 34 67 Joint displacements forjoints 33 and 56, and element stresses
and 67, for load case 101, is obtained with the help

for

elements 34

ofthe above commands.

FINISH

tThtschaptef

ts

adopted from staad Pro Help

Chapter Four

Application

Application
4-1 General This chapter deals with comparative study for concrete ground tanks with
constant dimensions and propenies ofmaterials. The tank is of6x4x4m and

of 0.2m thickness as shown in Fig.(4-l) the properties ofconcrete are:

compressive strength (Fcr=30), Modulus of elasticity (E=21.71 85) and Poisson's ratio (

v:0.l7).

The analysis is focused on the longest wall with

6x4 dimension. The connections between wall-wall and wall-slab are

considered as either fixed or pinned. For both cases displacements, stresses
and bendins moments calculations are done.

-T_
I

t
I

\,

/ a--

-l ------l b

Fig.(4-l)Dimension of tank

laol t,

Chapter Four

Application

4-2 Fixed Support

Casel : mesh (l xl)m

Displacement in z-direction for ground tank with

lxl

mesh

&.

&.

&,

Table.(4.1) Maximum Displacement in z-direcrion Node No

8 7

Displacement (mm)
0.000 -0.645
-1.373

l8
25

-t.602
-1.632

chapter Four

Application

Case2 : mesh (0.5x0.5)

Displacement in z-direction for ground tank with 0.5x0.5 mesh

Node No

Displacement (mm)
0.000 -0.245 -0.722

t4

l3
33

46 59 72
85

-1.190
-1.541

-1.755 -1.861
-

98

1.910

lll

-1.959

Chapter Four

Application

Case3 : mesh (0.25x0.25)

Displacement in z-direction for ground tank with 0.5x0.5 mesh

....:. ...:::: ;. :..:

'...'

a h

[l
&

...;.. ;.:.. .. !:.:::. :.".':..;

:. ..:. .. ..'... !::. : ' :: :;: : 'i
$:

; .: ...;.. .. a .. :. : . .. : :.: :..:. *

t". +. :.
.

; .. .. .. ..

..

& & &

&

t.**"--

1:. :.. l-*a*n8. .

-*5t4"

i**r****on

Table.(4.3) Maximum Displacement in z-direction

Node No,

Displacement (mm)

Node No,

Displacement (mm)

L0

0.000 -0.077
-0.255

238
2.OJ

-1.7 13

l)
63

-1.811

288

-1.883 -1.933

88
113

-0.488 -0.741 -0.990

3ll
338

-1.970
-2.000

138
163

J6l
388 413

-1.219
-

-2.029 -2.065

188

1.418

213

-1.583

{,,

Chapter Four

Application

Displacement in z-direction for ground tank with fixed support

4.5

3.5
3

2.5
2
'1.5
1

+m0.25x0.25mesh +m0.5x0.5mesh --rFmlxl mesh

0.5
0

Fig.(4-2) Displacement in z-direction for ground tank with fixed support

rJJ_

Chapter Four

Application

r::' .:li
a:::
a::t-

r:T

::3

r;:
a.:

al;

l,;
a':

: r::

r,;

(A)

!ip
:::"

rl;
i::i
a:x

r:; !t::

a;; r;;: r;l;

(B) Fig.(4-3)Continued

rJ--

Chapter Four

Application

a:::
a:t
I

a:.'

rll.
r.;. r:;,

r;r;;:

!:;
r::

!;;

(c)
Fig.(4-3) Max Von Mis stress for ground tank with fixed support (A) mesh

lxlm

(B) mesh 0.5x0.5m

(C) mesh 0.25x0.25m

ll=#

Chapter Four

Application

I' r:i a"":

il:

t:t;:

rll .l; !_..

r;."

!;;,
a::,

::x

(A)
ti," .l:l rl.a .:: rl:: t -.. .:.: ..;, a: rll;
r;;.

:l'.
r,";

r:l:
a,;;

(B) Fig.(4-4)Continued

Chapter Four
l'il',

Application

rl-: r;l a:r'l::

r'1:'

r:1"

r l::

r-;
t:::
:r"l, r::l r;":: t,;

(c)
Fig.(4-4) Mx for ground tank with fixed support
(A) mesh

lxlm

(B) mesh 0.5x0.5m

(C) mesh 0.25x0.25m

s7

Chapter Four

Application

ff: t.^; r. -: al:l ::I r: .::'

rj,::

!lr'
a:".

r:;
t;;

(A)

ltl r;. t,;;

!
_-:

:::i:

!::: t -:t r:j:

!:l; "llI

(B)
Fig.(4-5)Continued

{s8

Chapter Four

Application

:li
it; t.;;
r l::

iit:
r-"::

:ii"
!;;; t*

(c)
Fig.(4-5) My for ground tank with fixed support (A) mesh

lxlm

(B) mesh 0.5x0.5m

(C) mesh 0.25x0.25m

Chapter Four 4-3 Pinned Support

Application

Casel : mesh (1x1)m

Table.(4.4)Displacement in z-direction for ground tank with 1x1 mesh Node No
8 7

Displacement (mm)
0.000 -1.598 -6.002

t8
1,)

-7.354
-8.396

Case2 : mesh (0.5x0.5)

Table.(4.5)Displacement in z-direction for ground tank with 0.5x0.5 mesh Node No

Displacement (mm)
0.000

t4
13

-2.084
-3.933 -5.455

JJ

46 59 72
85 98
111

-6.636
-7

.530

-8.224

-8.826

-9.450

t60ffi

Chapter Four
Case3 : mesh (0.25x0.25)

Application

Table.(4.6)Displacement in z-direction for ground tank with 0.5x0.5 mesh

Node No.

Displacemetrt (mm)

Node No.

Displacemetrt (mm)
-7

0.000
25 63 88
113

238
263 288
313

.299

-1.084

-2.129
-3.1 19

-8.1

l7

-8.463

-4.023
-4.848

338
363

-8.787
-9.

138
163

t0l

-5.584

388

-9.420
-9.7 57

188

-o.l t4
-6.803

213

Displacement in z-direction for ground tank with pinned support

4.5

4
3.5
3

2.5
2
'1.5
1

.99

-+m1x1mesh .+m0.5x0.5mesh

-a--

mu lSxu l5mesh

0.5
0

Fig.(4-6) displacement in z-direction for ground tank with pinned support

Chapter Four

Application

(A)

(B) Fig.(4-7) Continued

t,

Chapter Four

Application

rlI'
I:] r,:'

::" r;: t;:

r':l
I::; r;;. r;;; .,:,
.:1:

!;;
E:::

(c)
Fig.(4-7) Max Von Mis stress for ground tank with pined support

(A) mesh

lxlm

(B) mesh 0.5x0.5m

(C) mesh 0.25x0.25m

Chapter Four

Application

# a;;

r.;; r:.. r;. r.;;

i*: -::

!::
r;;

(A)

(B)
Fig.(4-8) Continued

Chapter Four

Application

lil r,.;
t.;
r";"

:::

rl.: t: rl::
r..;

!;i;

"ll' r::.
r;,"

r,;.

(c)
Fig.(4-8) MX for ground tank with pined support

(A) mesh

lxlm

(B) mesh 0.5x0.5m

(C) mesh 0.25x0.25m

(--

Chapter Four

Application

lx r;;
a::
r.;: r::: r:1

:::

t.". ..;" r.;. a,^, r.;; r;,;

!;;'

(A)

-:i'
ri:r:: r;::

!::i
t;:; a:: .t:.
r.;.

t^::

::i:

(B)
Fig.(4-9) Continued

{uu

Chapter Four

Application

(c)
Fig.(4-9) My for ground tank with pined support

(A) mesh

lxlm

(B) mesh 0.5x0.5m

(C)mesh 0.25x0.25m

l,

Chapter Flve

Conclusions and Recommendations

Conclusions and Recommendations

5-1 Conclusions Many methods are available to analyze tanks like finite element method, grillage method,....etc . This research has used a commercial package program, Staad
Pro 2004 . This program utilize finite element method in solving for different
t)4)es of structures. In this study a plate/shell element is used to model the side

walls oftanks. Due to the present study

comparison is made to determine the

behavior ofthe tank walls, specifically : displacements, stresses and bendins moments.

1- For finer mesh sizes is

seen that increase in displacements for fixed_

supported walls of (2lo/o) fot 0.25mthanthat of 1m mesh size,

respectively.
For pin-supported walls the ratio will become (14%) for 0.25m than that of mesh size, respectively.

lm

This proves that the design ofthe tanks using finite element method needs finer
meshes to satisry the requirement ofthe tank functionality
.

2-

The estimation ofstresses is highly affected by type ofboundary

conditions. Also the bending moments in both directions ( x,y induced in ) the walls will give much diflerent results.
The maximum von mises stresses for fixed supported walls are located at the mid-bottom edge ofthe wall, while the stresses axe located in the lower comers

ofthe wall, for pinned supported walls.

3- For both fixed

supported and pin supported walls the maximum moments

(Mx) located in mid-top edge, while the maximum moments (My) located

68rt

Chapter Five
at the center

Conclusions and Recommendations

ofthe wall.

4- The maximum moment (Mx) for pin supported is as larger as 2.g times that of fixed supported walls. The maximum moment (My) for pin support is as larger as 2.5 times that

of fi xed supported walls.

5-2 Recommendations

l-

to analyze underground and./or elevated tanks using the Staad pro

program.

Taking in consideration the effect offlexibility ofthe connection lines

between walls and slab.

J. Studying the behavior ofthe base slab on the ground.

4- Using another method like grillage method in the analysis.

FBSTRAcI
Water storage tanks are very popular structure and frequently used in the civil engineering. The most important issue is to satis$, the structural and

functional adequacy. In the design process, many methods are availabte to delermine the material and geomeffic requirement. This study has used staad pro. 2004 program which use finite element method. The platelshell element is used to model the walls ofthe tanks.
The finite element method solution depends largely on the mesh size. is noted that using finer mashes supported walls idealization more than (1m) mesh size.

It

will

lead to more accurate results. Fixed_ increase of(0.25 m) mesh size

will give a 2l%

For pin-supported walls idealization, the maximum moment (M") is as

large as 2.9 times that of fixed supported.

References

References

l-

Reinforced concrete structure by I.C. syal and A.K. Goal, lggg second edition, Page( l-6)

2- http://www.hacengineers.co.uk/pdfs/tankVTANKSFINAL.pdf.oom,
Page(7-11)

3- Bairagi, N.K., Plate Analysis, Khanna publishers, New Delhi, First

Edition, 1986, Page(I2- 19)

4-

Concrete liquid Retaining structures (1980) by J.Keith Green and philip

H. Perkins, Page(1 9-30)

*zo