DIRECT STIFFNESS METHOD FOR TRUSSES: - PowerPoint PPT Presentation

About This Presentation
Title:

DIRECT STIFFNESS METHOD FOR TRUSSES:

Description:

CHAPTER 3 DIRECT STIFFNESS METHOD FOR TRUSSES: 3.1 INTRODUCTION In the previous chapter the procedure for obtaining the structure stiffness matrix was discussed. – PowerPoint PPT presentation

Number of Views:2115
Avg rating:3.0/5.0
Slides: 55
Provided by: 32505
Category:

less

Transcript and Presenter's Notes

Title: DIRECT STIFFNESS METHOD FOR TRUSSES:


1
  • CHAPTER 3
  • DIRECT STIFFNESS METHOD FOR TRUSSES
  • 3.1 INTRODUCTION
  • In the previous chapter the procedure for
    obtaining the structure stiffness matrix was
    discussed. The structure stiffness matrix was
    established by the following equation
  • K TT kc T -----------(2.21)
  • However if a large and complicated structure is
    to be analyzed and if more force components are
    to be included for an element then the size of
    composite stiffness matrix kc and deformation
    transformation matrix T will be increased.
    Therefore the procedure outlined in the preceding
    chapter for formation of structure stiffness
    matrix appears to be inefficient, furthermore
    this procedure is not suitable to automatic
    computation.

2
  • In this chapter an alternative procedure
    called the direct stiffness method is
    introduced. This procedure provides the basis for
    most computer programs to analyze structures. In
    this method each individual element is treated as
    a structure and structure stiffness matrix is
    obtained for this element using the relationship
  • Km TTmkm Tm
    -----------(3.1)
  • Where
  • Km Structure stiffness matrix of an
    individual element.
  • Tm Deformation Transformation matrix of an
    individual element.
  • km Member Stiffness matrix of an individual
    element.
  • Total structure stiffness matrix can be obtained
    by superimposing the structure stiffness matrices
    of the individual elements..

3
  • As all members of a truss are not in
    the same direction i.e. inclination of the
    longitudinal axes of the elements varies,
    therefore stiffness matrices are to be
    transformed from element coordinate system to
    structure or global coordinate system. When the
    matrices for all the truss elements have been
    formed then adding or combining together the
    stiffness matrices of the individual elements can
    generate the structure stiffness matrix K for
    the entire structure, because of these
    considerations two systems of coordinates are
    required.
  • i) Local or member or element coordinate system
  • In this coordinate system x-axis is
    collinear with the longitudinal axis of the
    element or member. Element stiffness is
    calculated with respect to this axis. This system
    is illustrated in figure 3.1

4
(No Transcript)
5
ii) Structure or global coordinate system A
single coordinate system for the entire structure
is chosen, with respect to which stiffness of all
elements must be written. 3.2 PROCEDURE FOR THE
FORMATION OF TOTAL STRUCTURE STIFFNESS MATRIX FOR
AN ELEMENT USING DIRECT STIFFNESS METHOD
Following is the procedure for the
formation of structure stiffness matrix
i) Formation of the element stiffness matrix
using equation 2.16.
----------- (2.16) ii) Formation of the
deformation transformation matrix T for a
single element ?m Tm ?m
-----------(3.2)
6
where ?m Element or member deformation
matrix. ?m Structure deformation matrix
of an element or member Tm Element or
member deformation transformation
matrix. (iii) Formation of a structure stiffness
matrix K or an element using the
relation Km TTmkm Tm
-----------(3.1)
The above mentioned procedure is discussed in
detail in the subsequent discussion. 3.2.1 The
formation of element stiffness matrix in local
co-ordinates It has already been discussed in
the previous chapter. However it is to be noted
that for horizontal members the structure
stiffness matrix and element stiffness matrix are
identical because both member coordinate systems
and structure coordinate systems are identical.
But for inclined members deformation
transformation matrices are to be used because
member coordinate system and structure coordinate
system are different therefore their structure
stiffness matrix and element stiffness matrix
will also be different.
7
  • 3.2.2 The formation of deformation
    transformation matrix
  • As the main difference between the previously
    discussed method and direct stiffness method is
    the formation of the deformation transformation
    matrix. In this article deformation
    transformation matrix for a single element will
    be derived.
  • Before the formation of deformation
    transformation matrix following conventions are
    to be established in order to identify joints,
    members, element and structure deformations.
  • 1) The member is assigned a direction. An arrow
    is written along the member, with its head
    directed to the far end of the member.
  • 2) ?i, ?j, ?k, and ?l, are the x and y structure
    deformations at near and far (tail head) ends
    of the member as shown in figure 3.2. These are
    positive in the right and upward direction.
  • 3) ?r and ?s are the element deformations at near
    and far (tail head) ends of the member as shown
    in figure 3.2.
  • 4)The member axis (x-axis of member coordinate
    system) makes an angle ?x, with the x-axis of the
    structure coordinate system as shown in figure
    3.2.
  • 5)The member axis (x-axis of member coordinate
    system) makes an angle ?y, with the y-axis of the
    structure co-ordinate system as can be seen from
    figure 3.2.
  • 6) The cosines of these angles are used in
    subsequent discussion Letters l and m represent
    these respectively.

8

9
l Cos qx and m Cos qy l Cos qx
m Cos qy
The algebraic signs of ?s will be automatically
accounted for the members which are oriented in
other quadrants of X-Y plan.
In order to form deformation transformation
matrix, once again consider the member of a truss
shown in figure 3.1.
10
(No Transcript)
11
(No Transcript)
12
Following four cases are considered to form
deformation transformation matrix. Case-1
Introduction of horizontal deformation to the
structure ?i 1, while far end of the member is
hinged (restrained against movement). From the
geometry of figure 3.3(a) ?r ?i Cos qx 1.
Cos qx Cos qx l ?r l ----------
(3.3) ?s 0 ---------- (3.4) Case-2
Introduction of vertical deformation to the
structure ?j 1, while near end of the member is
hinged (restrained against movement). From the
geometry of figure 3.3 (b) ?r ?j Sin qx ?j
Cos qy 1. Cos qy m ?r m ----------
(3.5) ?s 0 ---------- (3.6) Case-3
Introduction of horizontal deformation to the
structure ?k 1, while near end of the member is
hinged (restrained against movement).From the
geometry of figure 3.3(c) ?s ?k Cos qx 1.
Cos qx Cos qx l ?r 0 ----------
(3.7) ?s l ---------- (3.8)
13
  • Case-4 Introduction of vertical deformation to
    the structure DL 1, while near end is hinged
  • From the geometry of figure 3.3(d)
  • ?L 1
  • ?r 0 ---------- (3.9)
  • ?s ? L
  • Cos qy m ---------- (3.10)
  • Effect of four structure deformations and two
    member deformations can be written as
  • ?r ?i Cos qx ?j Cos qy ?k .0 ? L .0
  • ?s ?i .0 ?j .0 ?k Cos qx ? L Cos qy

?r l.?i m.?j 0.?k 0. ? L -----------
(3.11) ?s 0.?i 0.?j l.?k m. ? l
----------- (3.12) Arranging equations
3.11 and 3.12 in matrix from
----------- (3.13)
14
Comparing this equation with the following
equation ?m Tm ?m ---------------(3.2
) After comparing equation (3.14) and (3.2)
following deformation transformation matrix is
obtained.
-
---------- (3.14) This matrix Tm transforms
four structure deformation (?i, ?j, ?k, ?L ) into
two element deformation (?r and ?s).
3.2.3 Formation of structure stiffness
matrix Structure Stiffness matrix of an
individual member is first to be transformed from
member to structure coordinates. This can be done
by using equation 3.1. K TTm km
Tm --------(3.1)
----- (3.15)
15
------(2.16)
--------------------------------(3.16)
16
  • 3.2.4 ASSEMBLING OF THE INDIVIDUAL STRUCTURE
    ELEMENT STIFFNESS MATRICES TO FORM TOTAL
    STRUCTURE STIFFNESS MATRIX
  • Combining the stiffness matrix of the individual
    members can generate the stiffness matrix of a
    structure However the combining process should be
    carried out by identifying the truss joint so
    that matrix elements associated at particular
    member stiffness matrices are combined. The
    procedure of formation of structure stiffness
    matrix is as follows
  • Step 1 Form the individual element stiffness
    matrices for each member.
  • Step 2 Form a square matrix, whose order should
    be equal to that of structure deformations.
  • Step 3 Place the elements of each individual
    element stiffness matrix framed into structure in
    the corresponding rows and columns of structure
    stiffness matrix of step-2.
  • Step 4 If more than one element are to be placed
    in the same location of the structure stiffness
    matrix then those elements will be added.

17
Following figure shows the above mentioned
process.
18
In the above matrix the element in the second
row and second column is k1k2 where k1 is for
member 1 and k2 is for member 2.This is because
both k1 and k2 have structure coordinates 2-2.
19
Illustrative Examples Regarding the Formation of
K Matrix
Example 3.1 Form the structure stiffness matrix
for the following
truss by direct stiffness method.
20
Member Length Near End X1 Y1 Near End X1 Y1 Far End X2, Y2 Far End X2, Y2 i j k L lCos?x X2-X1/ Length mCos?y Y2-Y1/ Length
1 5L/6 0 0 L/2 2L/3 5 6 1 2 0.6 0.8
2 5L/6 L/2 2L/3 L 0 1 2 3 4 0.6 -0.8
3 L L 0 0 0 3 4 5 6 -1 0
According to eq. (3.16) we get
21
(No Transcript)
22
(No Transcript)
23
Example 3.2 Form the structure stiffness matrix
of the following truss.
24
Member Near End Near End Far End Far End Length lCosqx mCosqy i j k L
Member X1 Y1 X2 Y2 Length lCosqx mCosqy i j k L
1 0 0 L L ?2L 0.707 0.707 7 8 1 2
2 L L L 0 L 0 -1 1 2 5 6
3 L L 2L 0 ?2L 0.707 -0.707 1 2 3 4
So
25
Now the matrix can be written as k2
So k k1 k2 k3
26
(No Transcript)
27
  • 3.4 Force transformation matrix
  • The axial force (wm) in the members of a truss
    can be found by using the relationship between
    member forces and member deformations equation
    (2.15) and between element and structure
    deformations equation (3.1)
  • wm km ?m ------------ (2.15)
  • ?m Tm ?m ------------ (3.3)
  • Substituting value of ?m from equation 3.1 into
    equation 2.15
  • wm km Tm ?m ------------ (3.18)
  • Substituting value of km and Tm from equation
    2.16 and 3.15 respectively

28
In above Equations, wr Force at near end
See figure below
ws Force at far end
As wr -ws so
-------------- (3.19)
  • Sign conventions
  • wr ws

29
  • If
  • wr is positive then member is in compression.
  • ws is negative then member is in compression.
  • wr is negative then member is in tension.
  • ws is positive then member is in tension.
  • 3.5 ANALYSIS OF TRUSSES USING DIRECT STIFFNESS
    METHOD
  • Basing on the derivations in the preceding
    sections of this chapter a truss can be
    completely analyzed. The analysis comprises of
    determining.
  • i) Joint deformations.
  • ii) Support reactions.
  • iii) Internal member forces.
  • As the first step in the analysis is the
    determination of unknown joint deformation. Using
    the equations can do this.
  • W K ?
  • The matrices W, K and ? can be divided in
    submatrices in the following form

----------- (3.20)
30
  • Where
  • Wk Known values of loads at joints.
  • Wu Unknown support reaction.
  • ?u Unknown joint deformation.
  • ?k Known deformations, generally zero due to
    support conditions.
  • K11, K12, K21, K22 are the
    sub-matrices of K
  • Expanding equation 3.20
  • Wk K11 ?u K12 ?k ----------- (3.21)
  • Wu K21 ?u K22 ?k ----------- (3.22)
  • If the supports do not move, then ?k 0
    therefore equation 3.21 3.22 can be written as
  • Wk K11 ?u ----------- (3.23)
  • Wu K21 ?u ----------- (3.24)
  • By pre-multiplying equation 3.23 by K11-1
    following equation is obtained
  • K11-1 WKK11-1
    K11DU
  • ?u K11-1 Wk
    ----------- (3.25)
  • Substituting value ?u from equation 3.25 into
    equation 3.24
  • Wu K21 K11-1 Wk
    ----------- (3.26)

31
  • Using equation 3.25, 3.26 and 3.19, joint
    deformations, support reactions and internal
    member forces can be determined respectively.
  • As this method does not depend upon degree of
    indeterminacy so it can be used for determinate
    as well as indeterminate structures.
  • Using the basic concepts as discussed in the
    previous pages, following are the necessary steps
    for the analysis of the truss using stiffness
    method.
  • 1-Identify the separate elements of the structure
    numerically and specify near end and far end of
    the member by directing an arrow along the length
    of the member with head directed to the far end
    as shown in the fig. 3.2
  • 2-Establish the x,y structure co-ordinate
    system. Origin be located at one of the joints.
    Identify all nodal co-ordinates by numbers and
    specify two different numbers for each joint (one
    for x and one for y). First number the joints
    with unknown displacements.
  • 3-Form structure stiffness matrix for each
    element using equation 3.16.
  • 4-Form the total structure stiffness matrix by
    superposition of the element stiffness matrices.
  • 5-Get values of unknown displacements using
    equation 3.25.
  • 6-Determine support reaction using equation 3.26.
  • 7-Compute element or member forces using equation
    3.19.

32
3.6- Illustrative Examples Regarding Complete
Analysis of Trusses Example 3.3 Solve truss in
example 3.1 to find member forces.
33
The K matrix for the truss as formed in example
3.1 is
Using the relation W K ? we get
34
From equation (3.25) ?u K11-1 Wk
35
Using the relation Wu K21 ?u
Now the member forces can be found using the
relation
36
kips
kips
kips
37
Final Results sketch
38
Example 3.4
Solve the truss in example 3.2 to find member
forces.
39
The K matrix for the truss as formed in example
3.2 is as follows
Now as W k?
40
As ?k 0 or ?3, ?4, ?5, ?6, ?7,?8, all are zero
due to supports and using
Wk k11 ?u we get
Which gives
41
W1 0 W2 P Also using Wu k21 ?u we
have
Or

Or
42
Or
Or
Now Wu K21 ?u
43
W3 0.207 P W4 -0.207 P W5 0 W6 -0.5858
P W7 -0.207 P W8 -0.207 P
Now as
So
44
So, for member 1
Or ws 0.707 x 0.5858 P w1 0.414 P Similarly,
for member 2
Or w2 1 x 0.5858 P Or w2 0.5858 P
45
For member 3
Or w3 0.707 x 0.5858 P w3 0.414 P
46
w1 0.414 P w2 0.586 P w3 0.414 P
47
Example 3.5 Analyze the truss shown in the
figure.

A and E are constant for each member.
48
Member Length l m i j k l
1 10 -0.6 -0.8 7 8 1 2
2 11.34 -0.7071 0.7071 1 2 3 4
3 8 1 0 3 4 5 6
4 6 1 0 5 6 7 8
5 8 0 1 1 2 5 6
Using the properties given in the above table we
can find the structures stiffness matrices for
each element as follows.
49
(No Transcript)
50
(No Transcript)
51
Using equation K K1K2K3K4K5 we
get the structure stiffness matrix of (8x8)
dimensions. Partitioning this matrix with respect
to known and unknown deformations we get K11
and K12 portions as follows.
52

Using equation DuK111Wk we get the
Putting the above value in equation
WuK21Du
53
Element forces can be calculated using equation
as follows.
54
(No Transcript)
Write a Comment
User Comments (0)
About PowerShow.com