Professional Documents
Culture Documents
He has been active in teaching and research in structural analysis and mechanics
ANALYSIS OF
STRUCTURAL MEMBER
SYSTEMS
JEROME J. CONNOR
Massachusetts Institute of Technology
THE RONALD
Copyright
1976 by
Preface
With the development over the past decade of computer-based analysis methods, the teaching of structural analysis subjects has been revolutionized. The traditional division between structural analysis and structural mechanics became no longer necessary, and instead of teaching a preponderance of solution details it is now possible to focus on the underlying theory. What has been done here is to integrate analysis and mechanics in a systematic presentation which includes the mechanics of a member, the matrix formulation of the equations for a system of members, and solution techniques. The three fundamental steps in formulating a problem in solid mechanics. enforcing equilibrium, relating deformations and displacements, and relating forces and deformationsform the basis of the development, and the central theme is to establish the equations for each step and then discuss how the complete set of equations is solved. In this way, a reader obtains a more unified view of a problem, sees more clearly where the various simplifying assumptions are introduced, and is better prepared to extend the theory. The chapters of Part I contain the relevant topics for an essential background in linear algebra, differential and matrix transformations. Collecting this material in the first part of the book is convenient for the continuity of the mathematics presentation as well as for the continuity in the following development. Part II treats the analysis of an ideal truss. The governing equations for
small strain but arbitrary displacement are established and then cast into matrix form. Next, we deduce the principles of virtual displacements and
virtual forces by manipulating the governing equations, introduce a criterion for evaluating the stability of an equilibrium position, and interpret the governing equations as stationary requirements for certain variational principles. These concepts are essential for an appreciation of the solution schemes described in the following two chapters. Part III is concerned with the behavior of an isolated member. For completeness, first are presented the governing equations for a deformable elastic solid allowing for arbitrary displacements, the continuous form of the principles of virtual displacements and virtual forces, and the stability criterion. Unrestrained torsion-flexure of a prismatic member is examined in detail and then an approximate engineering theory is developed. We move on to restrained torsion-flexure of a prismatic member, discussing various approaches for including warping restraint and illustrating its influence for thin-walled
iii
PREFACE
and closed sections. The concluding chapters treat the behavior of planar and arbitrary curved members. How one assembles and solves the governing equations for a member sysopen
tern is discussed in Part IV. First, the direct stiffness method is outlined; then a general formulation of the governing equations is described. Geometrically nonlinear behavior is considered in the last chapter, which discusses member force-displacement relations, including torsional-flexural
coupling, solution schemes, and linearized stability analysis. The objective has been a text suitable for the teaching of modern structural member system analysis, and what is offered is an outgrowth of lecture notes
developed in recent years at the Massachusetts Institute of Technology. To the many students who have provided the occasion of that development, I am deeply appreciative. Particular thanks go to Mrs. Jane Malinofsky for her patience in typing the manuscript, and to Professor Charles Miller for his
encouragement.
JEROME J. CONNOR
Contents
IMATHEMATICAL PRELIMINARiES
1
18 19
Definition of a Matrix Equality, Addition, and Subtraction of Matrices Matrix Multiplication Transpose of a Matrix Special Square Matrices Operations on Partitioned Matrices Definition and Properties of a Determinant Cofactor Expansion Formula
3 5 5
8
10
12 16 19
21
Cramer's Rule 110 Adjoint and Inverse Matrices 111 Elementary Operations on a Matrix 112 Rank of a Matrix 113 Solvability of Linear Algebraic Equations
2
22 24 27 30
46
46 48 52
55 57
22 23 24
25
Introduction Second-Order Characteristic-Value Problem Similarity and Orthogonal Transformations The nth-Order Symmetrical Characteristic-Value Problem Quadratic Forms
66
66
71
32
33
Relative Extrema for a Function of One Variable Relative Extrema for a Function of n Independent Variables Lagrange Multipliers
75 81
81
42
82
CONTENTS
43 44 45
46
Unit Tangent Vector Principal Normal and Binormal Vectors Curvature, Torsion, and the Frenet Equations Summary of the Geometrical Relations for a Space
Curve
85 86 88
91
47 48
Local Reference Frame for a Member Element Curvilinear Coordinates for a Member Element
92 94
100
100 103 109
52 53
General
62
63
64
65
66 67
68 69
ElongationJoint Displacement Relation for a Bar General ElongationJoint Displacement Relation Force-Elongation Relation for a Bar General Bar ForceJoint Displacement Relation Joint Force-Equilibrium Equations Introduction of Displacement Restraints; Governing Equations Arbitrary Restraint Direction Initial Instability
152
152 153 159 162
165 169
General
72 73 74
Principle of Virtual Displacements Principle of Virtual Forces Strain Energy; Principle of Stationary Potential
Energy
75
76
178
178 178
General
82
83
180
CONTENTS
84
85
191
200
General
210
211
92 93
94
Governing EquationsAlgebraic Approach Governing EquationsVariational Approach Comparison of the Force and Mesh Methods
216 217
229
229 230 232 240 248
General
102
103 104
105 106
107
Summation Convention; Cartesian Tensors Analysis of Deformation; Cartesian Strains Analysis of Stress Elastic Stress-Strain Relations Principle of Virtual Displacements; Principle of Stationary Potential Energy; Classical Stability Criteria Principle of Virtual Forces; Principle of Stationary Complementary Energy
253
257
11
271
271
115 116
117
Introduction and Notation The Pure-Torsion Problem Approximate Solution of the Torsion Problem for Thin-Walled Open Cross Sections Approximate Solution of the Torsion Problem for Thin-Walled Closed Cross Sections Torsion-Flexure with Unrestrained Warping Exact Flexural Shear Stress Distribution for a Rectangular Cross Section Engineering Theory of Flexural Shear Stress Distribution in Thin-Walled Cross Sections
273
281
306
12
330
330
331
122
CONTENTS
123
Force-Displacement Relations; Principle of Virtual Forces Summary of the Governing Equations Displacement Method of SolutionPrismatic Member Force Method of Solution
13
371
132 133
134
135 136
137 138
139
371 Introduction Displacement Expansions; Equilibrium Equations 372 Force-Displacement RelationsDisplacement Model 375 Solution for Restrained TorsionDisplacement Model 379 Force-Displacement RelationsMixed Formulation 383 Solution for Restrained TorsionMixed Formulation 389 Application to Thin-Walled Open Cross. Sections -395 405 Application to Thin-Walled Closed Cross Sections Governing EquationsGeometrically Nonlinear Restrained Torsion 414
14
425
425 427
142 143
144
145
146
147
148
Introduction; Geometrical Relations Force-Equilibrium Equations Force-Displacement Relations; Principle of Virtual Forces Force-Displacement RelationsDisplacement Expansion Approach; Principle of Virtual Displacements Cartesian Formulation Displacement Method of SolutionCircular Member Force Method of Solution Numerical Integration Procedures
429
15
485
485 488
490 .493 499 507
511
152 153
154 155
156 157
158
Introduction; Geometrical Relations Force-Equilibrium Equations Force-Displacement RelationsNegligible Warping Restraint; Principle of Virtual Forces Displacement MethodCircular Planar Member Force MethodExamples Restrained Warping Formulation Member Force-Displacement RelationsComplete End Restraint Generation of Member Matrices
517
CONTENTS
Member MatricesPrismatic Member 1510 Member MatricesThin Planar Circular Member 1511 Flexibility MatrixCircular Helix 1512 Member Force-Displacement RelationsPartial End Restraint
159
520 524
531
535
545
545 546 547 548
Introduction Member Force-Displacement Relations System Equilibrium Equations Introduction of Joint Displacement Restraints
17
554
554 555 557 559
560 562 565 567 570 573
172
173
174 175
Governing Equations Network Formulation 177 Displacement Method 178 Force Method 179 Variational Principles 1710 Introduction of Member Deformation Constraints
176
18
585
585 585
591
Introduction Member EquationsPlanar Deformation Member EquationsArbitrary Deformation Solution Techniques; Stability Analysis
597
Index
605
Part I
MATHEMATICAL PRELIMINARIES
DEFINITION OF A MATRIX
ami,
a two-dimensional array, the first subscript defines the row location of an element and the second subscript its column location. A two-dimensional array having ,n rows and n columns is called a matrix of order m by n if certain arithmetic operations (addition, subtraction, multiplication) associated with it are defined. The array is usually enclosed in square brackets and written as*
a11 a21
a,,,1
a12 a22
am2
a1,,
a2,,
a,,,,,
Note
that the first term in the order pertains to the number of rows and the
second term to the nuiber of columns. For convenience, we refer to the order of a matrix as simply m x n rather than of order m by n.
*
CHAP. 1
A matrix having only one row is called a row matrix. Similarly, a matrix having only one column is called a column matrix or column vector.* Braces
instead ofbrackets are commonly used to denote a column matrix and the column subscript is eliminated. Also, the elements are arranged horizontally
instead of vertically, to save space. The various column-matrix notations are:
C11
C21
C1
C2
{c1, c2,.
{c1}
=c
If the number of rows and the number of columns are equal, the matrix is said to be square. (Special types of square matrices are discussed in a later section.) Finally, if all the elements are zero, the matrix is called a null matrix, and is represented by 0 (boldface, as in the previous case). Example
3
11
x 4 Matrix
4
3
21
7
1
8
1
1 x 3 Row Matrix
[3
4
2]
4Jor{3,4,2}
Square Matrix
5
[2
2 x 2 Null Matrix
[0 [o
0
* This is the mathematical definition of a vector. In mechanics, a vector is defined as a quantity having both magnitude and direction. We will denote a mechanics vector quantity, such as force or moment, by means of an italic letter topped by an arrow, e.g., F. A knowledge of Vector algebra is assumed in this text. For a review, see Ref. 2 (at end of chapter, preceding Problems).
MATRIX MULTIPLICATION
Two matrices, a and b, are equal if they are of the same order and if corresponding elements are equal:
a=
when
a=b
corresponds to mn equations:
= =
1,
2,. .
,m
,
1,
2,.. .
Addition and subtraction operations are defined only for matrices of the same
= =
+
bLJ]
For example, if
[1
then
ii
d
[1
[0 b=[3
1
1
i
0
1
3
and
[1
2
1
a+b=b+a
a+(b+c)=(a+b)+c
13.
(16)
(17)
MATRIX MULTIPLICATION
The product of a scalar k and a matrix a is defined to be the matrix in which each element of a is multiplied by k. For example, if
k=5
then
CHAP. 1
ka = ak = {ka11]
+ a12x2 +
a21x1 + a22x2 +
l2miXi
+ +
+
C1
C2
am2x2
i=
1, 2, .
. ,rn
where k is a dummy index. Using column matrix notation, (19) takes the form
i=
Now, we write (19) as a matrix product:
{c1}
i= 1,2,..,,rn
1,2
(111)
Since (110) and (1Il) must be equivalent, it follows that the definition equation for a matrix multiplied by a column matrix is
ax =
ulkxk}
j = 1, 2,. .
,m
This product is defined only when the column order of a is equal to the row
order of x. The result is a column matrix, the row order of which is equal to that of a. In general, if a is of order r x s, and x of order s x 1, the product ax is of orderr x 1.
Example
12
1
a=
11
2 x={3}
-4j
1(1)(2) + (1)(3)
4
+ (3)(3)
SEC. 13.
MATRIX MULTIPLICATION
We consider next the product of two matrices. This product is associated with a linear transformation of variables. Suppose that the n original variables x1, x2,. . . ,x,, in (19) are expressed as a linear combination of s new variables
Y1,Y2, . . . ,ys:
Xk =
1=
k=
1,
2,. .
,n
(113)
1,
2,. .
,m
k=i
the transformed equations take the form
i = 1,2 j
in
(114)
=
Noting (112), we can write (115) as
1,2,.. .,
py =
where p is in x .s and y is S x 1. Now, we also express the transformation of variables, in matrix form,
x = by
which defines
aby=c
and requiring (116) and (118) to be equivalent, results in the following definition equation for the product, ab:
=
ab
[bkJ] = [pt,]
1,2,.
. .
,n
This product is defined only when the column order of a is equal to the row order of b. In general, if a is of order r x n, and b of order n x q, the product ab is of order r x q. The element at the ith row and jth column of the product is obtained by multiplying corresponding elements in the ith row of the first matrix and the jth column of the second matrix.
CHAP. 1
Example 13
(1)(1) + (0)(O)
(O)(1) + (2)(O)
(IXI) + (O)(1)
(O)(1) + (2)(l)
(1)(O) +
1)
(1)( 1) + (01(3)
(1)(1) + (1)(3)
(0)(1)
+ (2)(3)
[+1
+1
0
ab=J_1
+4
+2 2 +6
If the product ab is defined, a and b are said to be confbrmable in the order stated. One should note that a and b will be conformable in either order only when a is in x n and b is n x in. In the previous example, a and b are conformable but b and a are not since the product ha is not defined. When the relevant products are defined, multiplication of matrices is associative,
(ab)c
(120)
a(b + c) = ab + ac (b + c)a = ha + Ca
Therefore, in multiplying b by a, one should distinguish preinultiplication, ab, from postrnultiplication ha. For example, if a and b are square matrices of order 2, the products are
[a11 [a21
[b11 [b21
b121
b22j
[aitbji
+ a12b21
a11b12 + a12b22
a21b12 + a22b22
[a21b11 + a22b21
aizl
a22]
[bjjaj1 + b12a21
[b21a11 + b22a21
b11a12 + b12a22
b21a12 + b22a22
When ab
ha,
14.
TRANSPOSE OF A MATRIX
is defined as the matrix obtained from a by The transpose of a = interchanging rows and columns. We shall indicate the transpose of a by
SEC. 14
TRANSPOSE OF A MATRIX
aT
{a79]:
a11
a12
a1, a2,
(123)
021
a22
=
amj
am2
a,,,
a21
= [a79] =
012
022
am2
a,,,
The element, a79, at the ith row and jth column of aT, where now i varies from 1 to n and j from 1 to m, is given by
a79 =
(124)
where
is the element at the jth row and ith column of a. For example,
[3
2
T
1
a =[2
r3
7
1
Since the transpose of a column matrix is a row matrix, an alternate notation for a row matrix is
[a1, a2
a,] =
(125)
We consider next the transpose matrix associated with the product of two
matrices. Let
m
Ilukbkf
1 .1
(b)
where now I =
write (c) as
1,
s and j = 1, 2,. .
p79 =
k1
2,.
k1
=
(ab)T
j=
(d)
bTaT
Equation (126) states that the transpose of a product is the product of the
10
CHAP. 1
Example
14
ab =
Alternatively,
aT
13 6
(ab)T = [4
13
6]
= [2
=
1]
1]
(ab)T = bTaT = [2
= [4
13
6]
15.
If the numbers of rows and of columns are equal, the matrix is said to be square and of order n, where n is the number of rows. The elements (i = 1, 2,. .. , n) lie on the principal diagonal. If all the elements except the principal-diagonal
elements are zero, the matrix is called a diagonal matrix. We will use d for
diagonal matrices. If the elements of a diagonal matrix are all unity, the diagonal
matrix is referred to as a unit matrix. A unit matrix is usually indicated by where n is the order of the matrix.
Example
15
Square
Matrix, Order 2
[1
7
[2 [o
Unit Matrix, Order 2
0
5
0
3
12[
0
I
LO
SEC. 15.
We
oij=0
+1
ij
i,j = 1, 2
n
(128)
(129)
d=
where d1, d2,. .
(130)
are the principal elements. If the principal diagonal elements . , are all equal to k, the matrix reduces to
=
and is called a scalar matrix.
(131)
Let a be of order rn x n. One can easily show that multiplication of a by a conformable unit matrix does not change a:
a
Ima = a
(132)
A unit matrix is commutative with any square matrix of the same order.
Similarly, two diagonal niatrices of order n are commutative and the product is a diagonal matrix of order a. Premultiplication of a by a conformable diagonal matrix d multiplies the ith row of a by and postmultiplication multiplies the jth column by
Example
16
[2 [o
01[3
01[2
01
[6
ij [2 7] [2 7
01
[3 11[2 [2
A square matrix a for which = property that a = If
7j[0 _1j[4
=
(i
[6
'
7
is called symmetrical and has the j) and the principal diagonal elements all equal zero, the matrix is said to be skew-symmetrical. In this case, aT = a. Any square matrix can be reduced to the sum of a symmetrical matrix and a skew-symmetrical matrix:
a=b+c
= = +
(1-33)
12
CHAP. 1
The product of two symmetrical matrices is symmetrical only when the matrices
are commutative.* Finally, one can easily show that products of the type
(aTa) (aaT)
(aTba)
A square matrix having zero elements to the left (right) of the principal diagonal is called an upper (lower) triangular matrix. Examples are:
Upper Triangular Matrix
2.
The transpose of an upper triangular matrix is a lower triangular matrix and vice versa. The product of two triangular matrices of like structure is a triangular matrix of the same structure.
[a11
[a21
0 1[b11
I
b22j
I=
[aijbij
[a21b11 + a22b21
a22b22
1-6.
Operations on a matrix of high order can be simplified by considering the matrix to be divided into smaller matrices, called .subina.trices or cells. The partitioning is usually indicated by dashed lines. A matrix can be partitioned in a number of ways. For example,
a11
012
a22
a32
0131
023
a11
a12
032
013
a33
a11
a12 a32
a13 a33
a21
031
= a1
031
a33J
a31
Note that the partition lines are always straight and extend across the entire matrix. To reduce the amount of writing, the submatrices are represented by
*
SEC. 16.
a
single symbol. We will use upper case letters to denote the submatrices whenever possible and omit the partition lines.
Example 1-1
We represent
[au
a12 a22
a13 a23
a=Ia,i
as
or
a = [A11
[A21
Ia13
I
A12
A22
where
A11
= [a21
=
[a31
a121
A12 =
La23
A21
a32]
A22 = [a33]
If two matrices of the same order are identically partitioned, the rules of matrix addition are applicable to the submatrices. Let
[A11 [A23
A121
[B11
A22J
[823
8121 B22j
(134)
where BLJ and A13 are of the same order. The sum is
a
+b =
[A11 + 8fl
LA2I + B21
(1-35)
The rules of matrix multiplication are applicable to partitioned matrices provided that the partitioned matrices are conformable for multiplication. In general, two partitioned matrices are conformable for multiplication if the partitioning of the rows of the second matrix is identical to the partitioning of the columns of the first matrix. This restriction allows us to treat the various submatrices as single elements provided that we preserve the order of multiplication. Let a and b be two partitioned matrices:
a
b
[A131t
= [B1d
= 1, 2,.. I = 1,2
M
M
(136)
k= 1,2,...,S
C = ab = [CIk]
M
1
C
when
,,...,
.
ik
i 1,
14
CHAP. 1
As an
au
ab =
1221
a13
h1
a23
033
h2 b3
1233
Suppose we partition a with a vertical partition between the second and third
columns,
1211
1212
a13
a=
a21
a22
a23
a33
= [A11A12]
a31
a32
For the rules of matrix multiplication to be applicable to the submatrices of a, we must partition b with a horizontal partition between the second and third rows. Taking
= [A,1A12]
= A11B11 + A12B21
The conformability of two partitioned matrices does not depend on the horizontal partitioning of the first matrix or the vertical partitioning of the second matrix. To show this, we consider the product
a12
a13
b11
b12
1322
ab
121 1231
1222
a23
1233
a32
b31
b32
1221 1231
a22
r A11
=
a32
a33
Since the column order of A11 and A21 is equal to the row order of b, no partitioning of b is required. The product is
ab =
[A111
LA2ijb = [A21b
b12
[A11b
b=
b21
b22
= [811B12]
b31
In this case, since the row order of B11 and B12 is the same as the column
SEC. 16.
ab =
a[B11B12]
= [aBj1
aBi2]
To transpose a partitioned matrix, one first interchanges the off-diagonal submatrices and then transposes each submatrix. If
A11
A12 A22
Am2
A1,,
a=
then
A21 Arnt
AT1 AT
AT
AT1
AT
.
AT
AT
AT
A particular type of matrix encountered frequently is the quasi-diagonal matrix. This is a partitioned matrix whose diagonal submatrices are square of various orders, and whose off-diagonal submatrices are null matrices. An
example is
a11
a= 0
0
0 a22
a32
0
a33
= [Ai
A2]
where
A1 = [a11]
A2 = [a22
a32
a23]
a33
and 0 denotes a null matrix. The product of two quasi-diagonal matrices of like structure (corresponding diagonal submatrices are of the same order) is
a quasi-diagonal matrix of the same structure.
A1
...
B1
...
A1B1
...
are of the same order. A and We use the term quasi to distinguish between partitioned and unpartitioned matrices having the same form. For example, we call
(140)
16
Cl-lAP. 1
17.
The concept of a determinant was originally developed in connection with the solution of square systems of linear algebraic equations. To illustrate how this concept evolved, we consider the simple case of two equations:
a11x1 +
a21X1
= + a22x2 =
a12x2
C2
c1a21
+ c2a11
a21 a2
order square array (i,j 1, 2). The determinant of an array (or matrix) is usually indicated by enclosing the array (or matrix) with vertical lines:
a11
a21
a12
a22
= al =
a31a22
a12a21
We use the terms array and matrix interchangeably, since they are synonymous. Also, we refer to the determinant of an eth-order array as an nth-order determinant. It shou'd be noted that determinants are associated only with square arrays, that is, with square matrices. The determinant of a third-order array is defined as
a11
a12
a22
a13 a33
+a11a22a33
a21
a31
a23 = a12a21a33
+ a12a23a31
(142)
a32
+a13a21a32 a13a22a31
This number is the coefficient of x1, x2, and x3, obtained when the third-order system ax c is solved successively for x1, x2. and x3. Comparing (l41) and (142), we see that both expansions involve products which have the following properties:
1.
2.
Each product contains only one clement from any row or column and no element occurs twice in the same product. The products differ only in the column subscripts. The sign of a product depends on the order of the column subscripts,
e.g., +a11a22a33 and a11a23a32,
These properties are associated with the arrangement of the column subscripts
described using
discussed below.
A set of distinct integers is considered to be in natural order if each integer is followed only by larger integers. A rearrangement of the natural order is called a permutation of the set. For example, (1, 3, 5) is in natural order and
SEC. 17.
(1,5,3) is a permutation of(1, 3,5). If an integer is followed by a smaller integer, the pair is said to form an inversion. The number of inversions for a set is defined
as the sum of the inversions for each integer. As an illustration, we consider the set (3, 1, 4, 2). Working from left to right, the integer inversions are:
Integer
3
1
Inversions
(3, 1)(3, 2)
Total
2
4 2
0
1
0
3
This set has three inversions. A permutation is classified as even (odd) if the total number of inversions for the set is an even (odd) integer. According to this convention, (1, 2, 3) and (3, 1, 2) are even permutations and (1, 3, 2) is an odd permutation. Instead of cbunting the inversions, we can determine the number of integer interchanges required to rearrange the set in its natural order since an even (odd) number of interchanges corresponds to an even (odd) number of inversions. For example, (3,2, 1) has three inversions and requires one interchange. Working with interchanges rather than inversions is practical only when the set is small.
Referring back to (141) and (142), we see that each product is a permutation
of the set of column subscripts and the sign is negative the permutation is odd. The number of products is equal to the number of possible permutations of the column subscripts that can be formed. One can easily show that there are possible permutations for a set of n distinct integers. We let , n) and define , ce,,) be a permutation of the set (1, 2,.
. .
.
as
I
1
when when
.
. ,
is
a,,)
an even permutation
(143)
..
is an odd permutation
a12 a22
a1,, a2,,
1
(144)
where
, n).
Factorial n =
1)(n
2)
(2)(1).
18
CHAP. 1
Example 18
The permutations for n =
3
are
a1=1
cxi1
x23
1
a33 a32
=3
e123=+1
e132=1
=2
z1=2
a3=1
a32
a3=1
e231=+1
e312=+1
e321-1
a12 a22
a32
a13
a11a22a33 a11a23a32
+a13a21a32 a13a22a31
2.
3.
4.
5.
6.
If all elements of any row (or column) are zero, the determinant is zero. The value of the determinant is unchanged if the rows and columns are interchanged; that is, aT! = a!. If two successive rows (or two successive columns) are interchanged, the sign of the determinant is changed. If all elements of one row (or one column) are multiplied by a number k, the determinant is multiplied by k. If corresponding elements of two rows (or two columns) are equal or in a constant ratio, then the determinant is zero. If each element in one row (or one column) is expressed as the sum of two terms, then the determinant is equal to the sum of two determinants, in each of which one of the two terms is deleted in each element of that row (or column).
7.
If to the elements of any row (column) are added k times the corresponding elements of any other row (column), the determinant is
unchanged.
[a31
[a21
a22
The determinant is
a! = a11a22 a12a21
a!. We
SEC. 18.
a' =
[a21
a22
a21 =
Then
a! = Next,
ka11
a22 = ka12
a12(ka1j)
a11(kaj2)
let
a11
+ c11
al
According to property 6,
hi + ci
=
ci
a21
where
b11
ibi
This
a21
b12
a22
a22
result can be obtained by substituting for O.ii and a12 in (b). Finally, to illustrate property 7, we take
b12 = a12 + ka22 b21 = a21 b22 = a7,
Then,
ibi
a!
1-8.
in the square matrix, a, If the row and column containing an element, are deleted, the determinant of the remaining square array is called the minor of and is denoted by The cofactor of is related to denoted by the minor of by (145) = ( As an illustration, we take
a=
The values of
328
531
1
and
M23. =
M22 =
= 37
37
20
CHAP. 1
Cofactors occur naturally when (.1 44) is expanded9 in terms of the elements
of a row or column. This leads to the following expansion formula, called Laplace's expansion by cofactors or simply Laplace's expansion:
=
a1kAIk
akJAkJ
(1 46)
Equation (146) states that the determinant is equal to the sum of the products of the elements of any single row or column by their cofactors. Since the determinant is zero if two rows or columns are identical, if follows that
k1
k
(147)
0 s
The above identities are used to establish Cramer's rule in the following section.
Example
(1) a11
(121
19
We apply (146) to a third-order array and expand with respect to the first row:
a12
a23
a32
a13 a23
a33
a31
a22
023
a33
+
+
023
a31
+ 0j3(
a22
1) 035
(133
a32
a11(a22a33 a23a32)
a52(a21a33
+ a23a31) +
a53(a21a32
022035)
To illustrate (1 47), we take the cofactors for the first row and the elements of the second row:
a21(a22a33
a23a32)
+ a22(a21a33 + a23a31) +
a23(a21a32
a22a31)
(2) Suppose the array is triangular in form, for example, lower triangular. Expanding with respect to the first row, we have
0
0
0
033
a21
031
a22
= a11
(122
0
033
(a51)(a22a33)
a11a22a33
a32
032
Generalizing this result, we find that the determinant of a triangular matrix is equal to the product of the diagonal elements. This result is quite useful.
* See
f See Ref. 4, sect. 3 15, for a discussion of the general Laplace expansion method. The expansion in terms of cofactors for a iow Or a COlUmn is a special case of the general method.
SEC. 19.
CRAMER'S RULE
The evaluation of a determinant, using the definition equation (144) or the cofactor expansion formula (146) is quite tedious, particularly when the array
is large. A number of alternate and more efficient numerical procedures for evaluating determinants have been developed. These procedures are described
in References 913.
c='ab
and we want cJ. It can be shown* that the determinant of the product of two square matrices is equal to the product of the determinants:
ci =
a! hi
(148)
Whether we use (148) or first multiply a and b and then determine lab! depends
on the form and order of a and b. If they are diagonal or triangular, (148)
is quite efficient. t
Example
110
[1
31
r2
=
and
Ic!
5]
a! =
Alternatively,
c
hi
20
[ [11
29J
[1
31
cj =
20
a=[0
a! = 5 Determining c first, we obtain
5]
bi = 8
r2 0 b__[1
Ic! = +40
rs
121
= [5 20]
and
ci = +40
19.
CRAMER'S RULE
=
*
= 1, 2, .
, ii
(a)
22
CHAP. 1
we
k1
Now, the inner sum vanishes when r
j=1
k.
This
The expansion on the right side of (c) differs from the expansion
al
ajrAj.
only in that the rth column of a is replaced by c. Equation (c) leads to Cramer's rule, which can be stated as follows:
If jaf = 0, a is said to be singular. Whether a solution exists in this ease will depend on c. All we can conclude from Cramer's rule is that the solution, if it exists, will not be unique. Singular matrices and the question of solvability
are discussed in Sec. 1 13.
110.
We
equations in n unknowns,
i,j
can be expressed as
1
1, 2,..., n
1, 2,. . ., ii
(note
Equation (e) leads naturally to the definition of adjoint and inverse matrices.
SEC. 110.
23
We define the adjoint and inverse matrices for the square matrix a of order n as
adjoint a = Adj a =
inverse a =
(149)
a1
Adj a
(150)
Note that the inverse matrix is defined only for a nonsingular square matrix.
Example
111
a= 2
The matrix of cofactors is
123 412
3
1
5 1
0
10
10
+7
1
+5
5
Adja
1/5
0
10
10 +5 +7 1
+ 1/25 + 2/5
7/25
+7/25
-- Adj
a=
1/5
+2/5
+ 1/25
x=
Substituting for x in (a) and c in (d), we see that a1 has the property that
a1a = aa'
It follows that (1Si) is valid only when 0. Multiplication by the inverse matrix is analogous to division in ordinary algebra. If a is symmetrical,, then a is also symmetrical. To show this, we take the transpose of (15 1), and use the fact that a =. aT:
1
(a_la)T =
24
CHAP. 1
Premultiplication by a'
results in
a"'
and therefore a1 is also symmetrical. One can also show* that, for any
nonsingular square matrix, the inverse and transpose operations can be interchanged:
bT,_t =
(152)
We consider next the inverse matrix associated with the product of two square matrices. Let
c=
ab
where a and b are both of order n x n and nonsingular. Premultiplication and then b1 results in by
a'c = b
(b'a'')c =
It follows from the definition of the inverse matrix that
(ab)1 =
(153)
In general, the inverse of a multiple matrix product is equal to the product of the inverse matrices in reverse order. For example,
=
The determination of the inverse matrix using the definition equation (1 50)
2.
3.
The interchange of two rows or of two columns. The multiplication of the elements of a row or a column by a number other than zero. The addition, to the elements of a row or column, of k times the corresponding element of another row or column.
These operations can be effected by premultiplying (for row operation) or postmultiplying (for column operation) the matrix by an appropriate matrix, called an elementary operation matrix. We consider a matrix a of order x n. Suppose that we want to interchange
rowsj and k. Then, we premultiply a by an rn x in matrix obtained by modifying the mth-order unit matrix, I,,,, in the following way:
1. 2.
*
Interchange Interchange
and 5k and
SEC. 111.
25
We let e denote an elementary operation matrix. Then, ea represents the result of applying a set of elementary operations to the rows of a. Similarly,
ac represents the result of applying a set of elementary operations to thc columns of a. In general, we obtain e by applying the same operations to the conformable unit matrix. Since we start with a unit matrix and since the elementary operations, at most, change the value of the determinant by a nonzero scalar factor,* it follows that e will always be nonsingular.
Example
We
112
1/2
7
1
1/5
2
5
a=
We
first:
1.
2.
*
Add (3) times the first row to the second row. Add (2) times the first row to the third row.
properties of determinants (page 18).
See
26
CHAP. 1
These
100 0 201
1
1/2
1/5
0 0
11/2
2
7/5
27/5
1/2
1/5
1/2
1
1/5
0
0
2,'Il
0
0
1
0 0
11/2
2
7/5
14/55
27/5
27/5
Next, we add (2) times the second row to the third row:
1
0
1
1/2
1
1/5
1
14/55
269/55
Finally, we multiply the third row by 55/269. The complete set of operations is
100 010
0 0
0
1
0110
1
1
0
0
100
0
1
11/21/5
1
55/269
0
1/5
1/2
1
00
14/55 =b
1
This example illustrates the reduction of a square matrix to a matrix using elementary operations on rows, and is the basis for the Gauss elimination solution scheme (Refs. 9, 11, 13). We write the result as
ea =
0
e
0
0
6/11
2/11
+ 1870/2959
220/2959
55/269
We obtain e by applying successive operations, starting with a unit matrix. This is more convenient than listing and then multiplying the operation matrices for the various steps. The form of e after each step is listed below:
100 0 0 001
1
Initial
Step 1
3
1
100 0 201
Step 2
1
0 0
1
6/11
2
2/11
SEC. 112.
Step 3
1
RANK OF A MATRIX
27
0 2/11
4/11
0 0 0
[1
6/11
}
Step 4
0 2/11 0
6/11
+34/11
L+187o/2959
220/2959
0 55/269
Two matrices are said to be equivalent if one can be derived from the other by any finite number of elementary operations. Referring to Example 1 12, the matrices
1
1/2
7
1/5
2
5
1/2
1
and
21
00
1/5 14/55
1
where p and q are nonsinqular. This follows from the fact that the elementary operation matrices are nonsingular.
1-12.
RANK OF A MATRIX
The rank, r, of a matrix is defined as the order of the largest square array,
formed by deleting certain rows and columns, which has a nonvanishing deter-
minant. The concept of rank is quite important since, as we shall see in the next section, the solvability of a set of linear algebraic equations is dependent on the rank of certain matrices associated with the set.
a12 a22
ar2
aIr
01q
a21
an
azq
(155)
arr
0rq apq
We multiply the elements in rowj by (j 1, 2,. . . ,r) and subtract the result from the last row. This operation will not change the magnitude of Ar+t (see Sec. 17). In particular, we determine the constants such that the first r elements
28
CHAP. 1
012
022
a,2
0p2
(156)
a1,
a2r
apr
Equation (156) has a unique solution since the coefficient matrix is nonsingular. Then (1 55) reduces to
a11
012
022
a1,
(157)
a21
Ar+i
a,2
0
0rr
where
= apq
(158)
Orq]
follows that
apq
= [aiq. 02p
.,
Apr
:
0p1
(159)
012
022
r+2
m (160)
Equation (160) states that the last m r rows of a are linear combinations
of the first r rows. One can also show* that the last n r columns of a are
linear combinations of the first r columns.
Example
113
a=21 32
5
1234
7
12
14
SEC. 112.
RANK OF A MATRIX
29
We see that a is at least of rank 2 since the determinant associated with the first two rows
and columns is finite, Then, the first two rows are linearly independent. We consider the determinant of the third-order array consisting of columns 1, 2, and q:
1
2
1
ajq
a2q a3q
+ 223 = 5 22 = 7
we obtain
22a2q
3ajq
a3q
q = 3,4
Since a33 and (134 satisfy this requirement, we conclude that a is of rank 2. The rows are related by (third row) = + 3 (first row) + (second row)
One can show* that the elementary operations do not change the rank of a matrix. This fact can be used to dctcrmine the rank of a matrix. Suppose b
defined by (161) is obtained by applying elementary operations to a. We know
that band a have the same rank. It follows that a is of rank p. A matrix having the form of b is called an echelon matrix. When a is large, it is more efficient to reduce it to an echelon matrix rather than try to find the largest nonvanishing determinant:
a11
(121
(pxpt b12
II,
...
b2p B12
(161)
Example
114
[i a=)2
[5
First, we eliminate
2
1
3
3
2
12
12
4
8
0
0
* See
3 3 6
3 3
Prob. 140.
30
CHAP. 1
4
2
3 3 6
0
At this point, we see that r = 3. To obtain b, we multiply the second row by 1/3, the third row by 1/2, and interchange the third and fourth columns:
b= 0
1243 0010
1
(rnxn) (nxs) b c
(162)
One can show* that the rank of a cannot be greater than the minimum value
of r associated with b and C:
r(a)
(163)
[1/2
[1/2
+1/2 +1/2
01
1]
[0
It follows that a is of rank 1.
113.
0
1
[:: ::
Suppose a is of rank 2 and
a11
(1-64)
a21
* See
a22
(165)
Prob. 144.
SEC. 113.
If a is of rank 2, we can always renumber the rows and columns such that (165)
is satisfied. We partition a and x,
a
[a11 [a21
X1
a12
a22
a131
a23j
[A1
A2]
(166)
x1
1x2
and write (164) as A1X1 + A2X2 = c. Next, we transfer the term involving
(167)
0, it follows from Cramer's rule that (167) has a unique solution Since jA1j for X1. Finally, we can write the solution as
= Aj'(c
A2X2)
(168)
Since X2 is arbitrary, the system does not have a unique solution for a given c. The order of X2 is generally called the defect of the system. The defect for this
system is 1.
If a is of rank 1, the second row is a scalar multiple, say A, of the first row.
Multiplying the second equation in (164) by 1/A, we have
a11x1 +
C1
c2/A
(169)
2cr, the equations are inconsistent and no solution exists. Then, when a is of rank 1, (164) has a solution only if the rows of c are related in the same manner as the rows of a. If this condition is satisfied, the two equations in (169) are identical and one can be disregarded. Assuming that 0, the solution is
Tf c2
x1 =
(1/a11)(c1
a12x2
a13x3)
(170)
The defect of this system is 2. The procedure followed for the simple case of 2 equations in 3 unknowns is also applicable to the general case of in equations in n unknowns:
a11
(221
a12
a1,,
x1
C1
a is of rank in, there exists an mth order array which has a nonvanishing
determinant. We rearrange the columns such that the first in columns are
32
CHAP. 1
a12
az,,,
ai,m+1
a2m1
am,m1
Xm
az,,
= [
A1
(mxm)
A2
]
(172)
am2
amni
a,,,,
{x1
X2
Xm+i
.. x4 =
c A2X2
{ X1
(m
1)
((nrn)x 1)
X2 }
we write (171) as
A3X1
(173)
0, (173) can be solved for X1 in terms of c and X2. The defect of the set is n m, that is, the solution involves n m arbitrary constants represented by X2. Suppose a is of rank r where r < m. Then, a has r rows which contain an rth-order array having a nonvanishing determinant. The remaining m r rows are linear combinations of these r rows. For (171) to be consistent, that is, have a solution, the relations between the rows of c must be the same as those for a. The defect for this case is n r.
Since IA1I
Example
115
C1
(a)
= C3
+ 22C2
(c)
To show this, we multiply the first equation in (a) by these equations the third equation. Using (b), we obtain
0=
C3
I12C2
(d)
Unless the right-hand side vanishes, the equations are contradictory or inconsistent and no solution exists. When e 0, (c) is identically satisfied and we see that (a) has a nontrivial 0) only when r < 3. The general case is handled in the same manner.* solution (x
See
Prob. 145.
REFERENCES
33
In general, (1 71) can be solved when r < ,n if the relations between the rows of a and c are identical. We define the augmented matrix, ci, for (171) as
a11
a12
a2,,
C1
=
afl,2
[a
Cm
cJ
(174)
a,flfl
When the rows of a and c are related in the same way, the rank of tz is equal to
the rank of a. It follows that (171) has a solution only if the rank of the augmented matrix is equal to the rank of the coefficient matrix:
r(a)
(175)
Note that (175) is always satisfied when r(a) = m for arbitrary c. We can determine r(cz) and i(a) simultaneously using elementary operations on provided that we do not interchange the elements in the last column. The
where
> r(a) and has a nonvanishing element, is of rank r(a). If no solution exists. r(a), (171) contains r independent equations involving n unWhen
knowns. The remaining m r equations are linear combinations of these r equations and can be disregarded. Thus, the problem reduces to first finding r@) and then solving a set of r independent equations in n unknowns. The complete problem can be efficiently handled by using the Gauss elimmation
procedure (Refs. 9, 11, 13).
REFERENCES
1.
2.
3.
4.
5.
6.
7, 8.
R. A., W. J. DUNCAN and A. R. cOLLAR: Elementary Matrices, cambridge University Press, London, 1963. THOMAS, G. B., JR.: Calculus and Analytical Geometry. Addison-Wesley Publishing Co., Reading, Mass., 1953. BODBWIG, E.: Matrix calculus, Interscience Publishers, New York, 1956. HOUN, F. E.: Elementary Matrix Algebra, Macmillan Co., New York, 1958. HADLEY, G.: Linear Algebra, Addison-Wesley Publishing Co., Reading, Mass., 1961. HOUSEHOLDER, A. S.: The Theory of Matrices in Numerical Analysis, Blaisdell, Waltham, Mass., 1964. NOBLE, B.: Applied Linear Algebra, Prentice-Hall, New York, 1969. HIL DEBRAND, F. B.: Methods of Applied Mathematics, Prentice-Hall, New York,
1952.
9.
Faddeeva, V. N.: Computational Methods of Linear Algebra, Dover Publications, New York, 1959.
34
10.
CHAP. 1
RALSTON, A. and H. S. WILF: Mathematical Methods for Digital Computers, Vol. 1, New York, 1960. 11. RALSTON, A. and H. S. WILF: Mathematical Methods for Digital Computers, Vol. 2, Wiley, New York, 1967.
Wiley,
PROBLEMS
11.
(a)
.3
6j[ 3
(c)
(d)
[i
[i
[
2152
4J 55
[3
(e)
l][4
3j [2
[4
11
3j
[i
[
[2
12.
{b1, b2,
b,j
(b)
{ai,
(c)
a2, .
[b1, b2
[a11
[a21
[c1
01
c2j
a12
[o
a22
PROBLEMS
35
(d)
[Cii
[021
13.
a121
0
C2
022j [0
Sl= a1+ a2 + 03 =
S2
+ b2 + b3
3
bk
can be written as
3
S1S2
I
1
Generalize this result for the sum of n elements. 14. Suppose the elements of a and b are functions of y. Let
cia
db
[dblk]
dy
[dv j
c
dy
[dy j
= ab
then
dc
dy
15.
=a+b
db dy
da dy
Consider the triple product, abc. When is this product defined? Let
p
= abc
What is the order of p? Determine Determine an expression for aT. case where c 16. Evaluate the following products:
(a)
F
for the
41 [1
[2
(b)
ij [2 sj
21 [5
[4
1
1
17.
where a is a square matrix. (a + b)(a + b) Show that the product of two symmetrical matrices is symmetrical
only when they are commutative. 18. Show that the following products are symmetrical:
(a)
aTa
(b)
aTba
(c)
bTaTcab
36
CHAP. 1
1-9. Evaluate the following matrix product, using the indicated submatrices:
1
51
110. Let c = ab. Show that the horizontal partitions of c correspond to those of a and the vertical partitions of c correspond to those of b. Hint: See
Eq. (137).
111. A matrix is said to be symmetrically partitioned if the locations of the row and column partitions coincide. For example,
a11 a21 a31 012
0131
a22 a32
a23 a33
a12
a13
033
a32
i,j =
1,
2,. ., N
.
have
c=
(a)
ab
If a and b are symmetrically partitioned, show that CJk, AJk, the same order. Illustrate for the case of one partition, e.g.,
a
are of
[A11
[A2,
A12 A22
(b)
113.
Suppose we symmetrically partition c. What restrictions are placed on the partitions of a and b? Does it follow that we must also partition a and b symmetrically? Hint: See Prob. 110. Consider the triple product,
C=
a symmetric rth-order square matrix and a is of order r x ii. Suppose we symmetrically partition c. The order of the partitioned matrices are indicated in parentheses.
(pxp)
(pxq)
(nxn)
[C11
C12
[c21
(q><p)
c22
(qxg)
PROBLEMS
(a)
37
={A1
(r x q)
A2]
(b)
114.
Express
(j, k = 1,2) in terms of A1, A2, and b. Let d = [Di] be a quasi-diagonal matrix. Show that
da = bd =
when
the matrices are
[DJAJk] [BJkDk]
conformably
partitioned.
1iS.
following sets.
(a)
(b)
116. 117.
1a1122a1h3
The first subscripts in (a) are in natural order. We obtain (b) by rearranging (a)
such that the second subscripts are in natural order. For example, rearranging
e231 a12 a23 a31
= (2, 3, 1)
we obtain
e312a31a12a23
fi = (3, 1,2)
is
Show that if is an even permutation, (fl1, permutation. Using this result, show that
=
p
also an even
and, in general,
al =
118.
Suppose that
=
Then, (b) takes the form
Show that
(c)
= (a)
Generalize this result and establish that the sign of a determinant is reversed
when two rows are interchanged. 119. Consider the third-order determinant
a! =
38
CHAP. 1
Suppose
120. Suppose all the integers of a set are in natural order except for one integer, say n, which is located at position p. We can put the set in natural order by successively interchanging adjacent integers. For example,
121.
\
i=1
(
\j
=
t,J,
. .
=
Following this approach, establish Laplace's cofactor expansion formula for
0 0
o b12
b22
b2P
9
o
b11
0
(pxp) a
(pxn)
...i
a11
0n1
(un
(12,7
012
a22
bFa
= al
(nXn)
bni
123.
k12
0n2
d_[D1 [o (qxp)
By expressing d as
0
(qxq)
D2
[o
Iqj[0
D2
PROBLEMS
show
1100
2
3
d
Generalize for
d
124.
0021 0053
[A,
Let
g
[Gii
[G2,
= jG1
(pxp)
(pxq)
0
G22
(qxq)
Show that
G221
Generalize for a quasi-triangular matrix whose diagonal submatrices are square, of various orders. 125. Suppose we express a as the product of a lower triangular matrix, g, and an upper triangular matrix, b.
a11 a21
a12
g11
0
g22
0
0
b11
b12
b22
a22
'1n2
az,,
g21
?
0
1n2
Yflfl
bflfl
introduce symmetrical partitions after row (and column) p and write the product as
We
(pxp)
(pxq)
(pxp)
(pxq)
0
(pxp)
[A1,
LA21
(qxp)
A121
(qxq)
[G1,
(qxp)
1[Bi,
(qxp)
(pxq) B12
A22J [G2,
G22J[0
(qxq)
B22
(qxq)
Note that the diagonal submatrices of g and b are triangular in form. (a) Show that
= A.11 G11B12 = A12 G21B11 = A21
G21B12
(b)
+ G22B22 = A22
1G111
Show that
A111 =
BijI
and
al = 1G111 G22j 1B111 1B221
(c)
By taking p = 1, 2,.
,n
1,
40
CHAP. 1
aj1
aj2
The determinant of the array contained in the firstj rows and columns is called
the jth-order discriminant. 126. Does the following set of equations have a unique solution?
1
2
3 7
x1
X2
ii
x3
127.
a=
Does a
128.
123
1
3 7
11
exist?
bT,
(a)
[1
[3
(b)
[2 4
[i
(c)
[1
31[2
[3
(d)
2J[l
[2 O1[2
[o
Let
a12
3j[l
[A11
A12
A22
=
31
= [A2,
32
33
and
B32
[8,, 82,
where the order of BJk is the same as AJ1. Starting with the condition
aa'
13
PROBLEMS
41
determine the four matrix equations relating BJk and Aft (j, k
13!.
Use the results of Probs. 131 and 132 to find the inverse of
b
[B11
B12
[o
B22
where B11, B22 are square and nonsingular. Hint: write has
b 134.
[B11
01[1
0
1
[o
ij[o
a=
3
j[o B22
1121 1211
1
Determine the elementary row operation matrix which results in a21 = = Oand a11 022 033 = +1. 135. Let
(pxp)
(p'<q)
a31
a
where
[A11
A12
(qxq)
[A2,
A22
0. Show that the following elementary operations 0 and on the partitioned rows of a reduce a to a triangular matrix. Determine
1 [
L0
cj [A21 lqj [0
01
Aizi
AW
A22J
[o
Iq
AIA
\1
136.
2
1
3
3
a= 2
3
col2+coll col3*col2
42
CHAP. 1
Show that postmultiplication byIl(which is called a permutation matrix) results in the desired column rearrangement:
o
0
11
H=
01
i oj
rearranges the rows of a in the
(d) (e)
Show that 11TH 13. Generalize for the case where a is n x n. Show that
a!
2.. Show that a is of rank 1 when 137. Let a be of order 2 x n, where n the second row is a multiple of the first row. Also, show that when r = 1, the nth columns are multiples of the first column. second, third 138. Determine the rank of
(a)
1
5
3
4 10
2
3
(b)
1
2
1
2
1
2 3
139. Let a be of order in x ii and rank r. Show that a has n r columns which are linear combinations of r linear independent columns. Verify for
a= 2
1234
1
12
14
140. Using properties 3, 4, and 7 of determinants (see Sec. 17), deduce that the elementary operations do not change the rank of a matrix. For con-
venience, consider the first r rows and columns to be linearly independent. 141. Find the rank of a by reducing it to an echelon matrix.
PROBLEMS
142.
43
a2
C
:
a12
r b11b12
0.
cblk)
(j)2kj
(b)
Show that the second, third,. , nth columns of c are multiples of the first column and therefore r(c) 1. When will r(c) = 0? Suppose r(a) = 1 and a11 0. In this case, we can write
.
=
Show that the second, third
j = 2,3,...,m
1.
a12 a22
a,,,2
'
b11
b12 h22
b21
a,,1
h2 =
-
Let
-
.. ..
B1
{b11b21
A2
Am
Suppose r(a) = r4, r(b) rb. For convenience, we assume the first r0 rows of a and the first r1 columns of b are linearly independent. Then,
A3 =
p=
1
= ra + 1, ra + 2, -. , m
.
44
CHAP. 1
,
Bk
=
q
1
XkqBq
+ 1, Tb + 2, .
.
(a)
+ 2,. .
, in
the first rows. (b) Show that columns Tb + 1, of the first columns.
(c)
From (a) and (b), what can you conclude about an upper bound on
r(c)?
(d)
A1B1
A1B2 A2B2
AraB2
A1Br1,
A2
B2 [1 B ..
A2B1
rb]
A2Brb
Suppose ra
(e)
A1 is orthogonal
0
1
0 0
1
1 1
1/2
1
1/2
1
1
1
(f)
Suppose ra =
rb
s.
s.
Verify for
U
145.
X1
C1
022
C2
0R111
X,,
Let
=
Using (b), we write (a) as
1,
2,. .
, fl2
in = 1, 2 (c) Now, suppose a is of rank r and the first r rows are linearly independent. Then,
A3x =
A1
k = r + 1, r + 2
(a)
k=
p=1
+ 1, r + 2,.. .
, in
PROBLEMS
45
(b)
If m < n and r = m, the equations are consistent for an arbibrary c. Is this also true when rn > n and r = n? Illustrate for
[i [i
and
1
2
11
4]
1
1
X2
fc1
C1
1
1
c2
c3
146.
x1 + X2 + 2x3 +
2x4
2x1+x2+3x3+2x4=6 3x1-l-4x2+2x3+x4=9
7x1 +7x2+9.x3+7x4=23
Determine whether the above system is consistent using elementary operations on the augmented matrix. (b) Find the solution in terms of x4.
(a)
INTRODUCTION
(ajj
2)x1 + at2xz
(22)
The values of 2. for which nontrivial solutions of (2i) exist are called the
characteristic values of a. Also, the problem of finding the characteristic values and corresponding nontrivial solutions of (2i) is referred to as a second-order characteristic-value problem,* problem occurs naturally in the free-vibration The
analysis of a linear system. We illustrate for the system shown in Fig. 21. The equations of motion for the case of no applied forces (the free-vibration
case) are
d2y2
+ k2(y2
k2(y2
0
y1)
m1
d2y
+ k1y1
Also called "eigenvalue" problem in some texts. The term "cigenvalue" is a hybrid of the
SEC. 21.
NTRODUCT$ON
47
iwt
y2zize
A
and substituting in (a) lead to the following set of algebraic equations relating the frequency, w, and the amplitudes, A1, A2:
We can transform (c) to a form similar to that of(21) by defining new amplitude measures,* 2co2
A2 =
and the final equations are
k2 k1+k2_ == 2 = A
rn1
k2
A1 +
'fl2
A2 = 2A2
characteristic values and corresponding nontrivial solutions of (e) are related to the natural frequencies and normal mode amplitudes by (d). Note that the coefficient matrix in (e) is symmetrical. This fact is quite significant,
The
Although the application to dynamics is quite important, our primary reason for considering the characteristic-value problem is that results obtained for the characteristic value problem provide the basis for the treatment of quadratic
* See Prob. 2i.
48
PROBLEMS
CHAP. 2
forms which are encountered in the determination of the relative extrema of a function (Chapter 3), the construction of variational principles (Chapter 7), and stability criteria (Chapters 7, 18). This discussion is restricted to the case where a is real. Reference 9 contains a definitive treatment of the underlying theory and computational procedures.
22.
2)x1 + a12x2
A)x2
0
0
a21x1 + (a22
when
a11
a21
A
(24)
are possible only if the determinant of the coefficient matrix vanishes, that is,
a12
a22).
(2-5)
Expanding (25) results in the following equation (usually called the characteristic equation) for 1:
22
(26)
We let
(27)
+ P2 =
(28)
The roots of (28) are the characteristic values of a. Denoting the roots by
22, the solution is
(29)
(a11
a22)2
+ 4(a12)2
Since this quantity is never negative, it follows that the characteristic values for a symmetrical second-order matrix are always real.
Example
2i
[2 a={2
=2 5=
The characteristic equation for this matrix is
22
P2 = (2)(5) (2)(2) = 6
72
+6=
SEC. 22.
49
Solving (a),
[1
131=0
A.1
where
i=
By definition, nontrivial solutions of (24) exist only when 2 = or 22. In what follows, we suppose the characteristic values are real. We consider
a12x2
=
=
0 0
a21x1 +
(a22
This follows from the fact that the coefficient matrix is singular.
(Oii
%1)(a22
a12a21
Since only one equation isindcpendcnt and there are two unknowns, the soluAssuming* tiOn is not unique. We define as the solution for A = 0, the solution of the first equation is that a12
xi') =
1)
C1
=
012
where
and take c1 such that = 1. This operation is called normalization, and the resulting column matrix, denoted by Q1, is referred to as the characteristic vector for
a12
(210)
=
By definition,
+
L
_2112
J
-
Q1Q1 = 1
(211)
if a12
50
CHARACTERISTIC-VALUE PROBLEMS
CHAP. 2
Since
we see that
A1Q1
A2,
aQ1 =
(212)
we obtain
(213)
where
=
Also,
aQ2
+
L
j
(214)
It remains to discuss the case where A1 = A2. If a is symmetrical, the characteristic values will be equal only when a11 a22 and a12 = c121 = 0. Equation (24) takes the form (a11 A)xi + (0)xz 0
(O)x1 + (a11
A)x2 = 0
These equations are linearly independent, and the two independent solutions
{c1, 0}
{0, c2}
= {+1,0}
Q2 = {0, +1} If a is not symmetrical, there is only one independent nontrivial solution when the characteristic values are equal. It is of interest to examine the product, QfQ2. From (210) and (213),
(215)
we have
Q1Q2
C1C2
/1 +
(a11
-
A1)(a11
2 a12
A2)
(a22 A1)
a11
we see that QrQ2 = 0. This result is also valid when the roots are equal. In general, QIQ2 0 when a is unsymmetrical. Two nth order column vectors U, V having the property that
and
UTV=VTU=0
(216)
are said to be orthogonal. Using this terminology, Q1 and Q2 are orthogonal for the symmetrical case.
SEC. 22.
Example 22
[2 a=[2
A1=+6
The equations for A =
A2=+1
+6 are
4x1 + 2x2 = 0
2x1
=0
=
Then,
and
=
2}
= 2c1
Repeating forA =
and
A2
= +1, we find
=
Q2
One can easily verify that
c2{1,
=
and
j = 1,2
Ii a=[t
8
3
The characteristic values and corresponding normalized solutions for this matrix are
1
Q2 =
{4,
l}
We see that
0. Actually,
QTQ2 =
[1
A1_+i
A2=i
We have included this example to illustrate the case where the characteristic values are
52
PROBLEMS
CHAP. 2
are
2x2
(1 + i)x2 = 0
Note that the second equation is (1 i) times the first equation. The general solution is
x(1t =
Repeating for
c1
2=
12, we find
c2
1-3-i}
Now, we take c2 = c1. When the roots are complex, 12 is the complex conjugate of We determine c1 such that Then, xt2> is the complex conjugate of
= I
Finally, the characteristic values and characteristic vectors are
21,2 =
Qt,2 =
In general, the characteristic values are complex conjugate quantities when the elements of a are real. Also, the corresponding characteristic vectors are complex conjugates.
23.
aQ1 = aQ2 =
We can write (a) as
22Q2
Q2] = EQ1
2j
Q2]
(b)
Now, we let
q=[Qj
=
(217)
We call q the normalized Column j of q contains the normalized solution for modal matrix* for a. With this notation, (b) takes the form
aq =
(218)
* This terminology has developed from dynamics, where the characteristic vectors define the normal modes of vibration for a discrete system.
SEC. 23.
53
We have shown that the characteristic vectors are always linearly indepen-
dent when a is symmetrical. They are also independent when a is unsymmetrical, provided that 0 except for the case where a is 11.2. Then, unsymmetrical and the characteristic values are equal. If 0, q1 exists and we can express (218) as
q'aq =
(219)
The matrix operation, p is arbitrary, is called a similarity transformation. Equation (219) states that the similarity transformation, q1( )q, reduces a to a diagonal matrix whose elements are the characteristic values
of a.
Also, by definition,
q = [QTJ [Qi
[Qfl
Q2]
qT
[1
0
1
[o
A square matrix, say p, having the property that is called an = orthogonal matrix and the transformation, pT( )p, is called an orthogonal transformation. Note that an orthogonal transformation is also a similarity transformation. Then, the modal matrix for a symmetrical matrix is orthogonal and we can write qTaq = (221)
Example
23
p'
[2
+6
2
5
Q2{2,-1}
[0
0
+1]
q=[Qi
54
PROBLEMS
CHAP. 2
We verify that qT
21 [1
21
1 [5
01
lj[2
1
aq
[2 21[1 [2
[1
21
1 [6
sj[o i
2
[1
ij
[12
aq
i[i
= 5L2
ij[o
21[6 01
21
21[6
[6
01
_-1][12 ii = [o
ij =
(2)
[1
8 3
= +5
Qi =
+1}
01
Lo
Since
ij
q
a is not symmetrical, qT
q
Actually,
-1
[
One can easily verify that
q
j
[5
01
[,15/6 [\/17/6
[1
01
Lo iJ
[1
q involves complex elements. Since the characteristic vectors are complex conjugates, they are linearly independent and q -' exists. We find q 1, using the definition equation for the inverse (Equation (150)):
SEC. 24
55
-1
[+i
0
01
24.
The nth order symmetrical characteristic-value problem involves determining the characteristic values and corresponding nontrivial solutions for
a11x1 + a12x2 + a12x1 + +
+ +
= Ax1
AX2
(222)
+
We can write (222) as
+
ax = AX
(223)
(a
In what follows, we suppose a is real. For (223) to have a nontrivial solution, the coefficient matrix must be singular.
a AI,4
0
(224)
+
where
+
+
= a11 + a22
and
(225)
-
is the sum of all the jth order minors that can be formed on the diagonal.* Letting 22,. . , denote the roots, and expressing the characteristic equation in factored form, we see that
.
2522
+ 2523 + ... +
(2-26)
We summarize below the theoretical results for the real symmetrical case. The proofs are too detailed to be included here (see References 1 and 9):
1.
2.
The characteristic values are all real. 22,. . , The normalized characteristic vectors Q1, Q2,. . , Q,,, are orthogonal:
. .
QTQJ =
i,j =
1,2
* Minors having a diagonal pivot (e.g.. delete the kth row and column). They are generally called principal minors.
56
3.
CHARACTERiSTIC-VALUE PROBLEMS
CHAP. 2
where
=
Example 24
52
0
.1
1
a= 2
0 using (225):
Since a is symmetrical, its characteristic values are all real. We first determine /3k, $2, f33,
5+
$2
/33
+1
+9
1826=0
and the approximate roots are
22
2x2
x3 = (3 2)x2
(l,t)x3=
x2
Solving the first and third equations for x1 and x3 in terms of x2, the general solution is
j=l,2,3
Finally, the modal matrix (to 2-place accuracy) is
+0.22
q = [Q1Q2Q3] = +0.50
+0.85
(2)
+0.51 +0.68
0.52
0.84
+0.54
0.10
a= 2
120
1
0
3
0 .0
SEC. 25.
QUADRATIC FORMS
213J
57
The expansion of Ia
0 is
223
Writing out ax = 2x, we have
(12)x1
2x1
23=i
+2x2
+(1 1)x2
(3 2)x3
=0
= =
0 0
(a)
When 2 =
3,
(a) reduces to
2x3
+2x2 =
2x2 = 0
(b)
(0)x3=0
We see from (b) that (a
213)
is of rank 1 when 2 =
3.
xI=c1
x2=cl
x3=cz
By specializing the constants, we can obtain two linearly independent solutions for the repeated root. Finally, the characteristic vectors for 22 = 3 are
Q2 = (0,01)
When 2 = 23 = 1, (a) reduces to
2x3 + 2x2
2x1 + 2x2
4x3
=0
are
=
o}
This example illustrates the case of a symmetrical matrix having two equal characteristic
values. The characteristic vectors corresponding to the repeated roots are linearly independent. This follows from the fact that a 213 is of rank I for the repeated roots.
F
a quadratic form in
+ 2a12x1x2 +
F as
F=[xix2]t[a12
ajal
58
CHARACTERISTIC-VALUE PROBLEMS
CHAP. 2
F=
where afk =
=
j=1
.
(227)
, x,,.
for]
for
If F =
a
positive definite a 0, we say that F is positive semidefinite. We define negative definite and negative semidefinite quadratic forms in a similar manner. A quadratic form is negative definite if F 0 for all x and F = 0 only when x = 0. The question as to whether a quadratic form
is positive definite is quite important. For example, we will show that an equilibrium position for a discrete system is stable when a certain quadratic
form is positive definite.
F=
b1
LXIX2
13
b2
x2
(228)
b1 >0
It is positive semidefinite when
b1
b2
>0
and at least one of the elements is zero. Now, to establish whether is positive definite, we first reduce a to a diagonal matrix by applying the transformation, q '()q, where q is the orthogonal normalized modal matrix for a. We write xTax = (xTq)(q_taq)(qlx)
=
Then, letting
y=
qTx
qy
(229)
F = xTax =
(230)
It follows that F is positive definite with respect to y when all the characteristic values of a are positive. But y is uniquely related to x and y = 0 only when x = 0. Therefore, F is also positive definite with respect to x. The problem of establishing whether xTax is positive definite consists in determining whether all the characteristic values of a are positive.
SEC. 25.
QUADRATIC FORMS
[a11
cz12
Laiz
a22
+
,t122 = 132
=aii -F a22
=
a11a22
aJ
132>0
are
equivalent to
,t2>0
Suppose we specify that
au
>
at . a11a22
>
Since a1 > 0, it follows from the second requirement in (d) that a22 > Therefore, (d) is equivalent to (b). We let
0.
= aj1j =
A2 =
Then, a is positive definite when
a11a12
a11
a12a12
at
(231)
132>0
(23 2)
A1>O
The quantities
A2>0
and are called the invariants and discriminants of a. The above criteria also apply for the case. That is, one can show that a is positive definite when all its invariants are greater than zero.
131 > 0
>
..
/3,,
>
(233)
where
is
can be expressed in terms of the discriminants. Let represent the determinant of the array consisting of the first j rows and columns.
a11
a12
aU
(234)
A=
The conditions,
A1 >
0
112
a22
a2J
A2 >
...
A,,
>
(235)
See
60
CHARACTERISTIC-VALUE PROBLEMS
CHAP. 2
Example 25
111
The discriminants are
122 123
$2(21)+(31)+(64)
/13
+5
= A3 = +1
1
1 1
2
3
Since A2 is negative
3),
p1ap
pTap
(236)
bT =
pTaTp
(237)
Now, b and a have the same characteristic values.* This follows from
(238)
Then, if a is positive definite, b is also positive definite. In general, the positive definite character of a matrix is preserved under an orthogonal transformation,
REFERENCES
1.
2. 3. 4. 5.
BODEWIG, E.: Matrix Calculus, Interscience Publishers, New York, 1956. SMiRNOV, V. I.: Linear Algebra, Addison-Wesley Publishing Co., Reading, Mass.,
1964.
TURNBULL, H. W., and A. C. AITKEN: An Introduction to the Theory of Canonical Matrices, Dover Publications, New York. HADLEY, G.: Linear Algebra, Addison-Wesley Publishing Co., Reading, Mass., 1961.
Prob. 25.
* See
PROBLEMS
6.
7.
8.
9.
10.
11.
12.
13.
CRANDALL, S. H.: Engineering Analysis, McGraw-Hill, New York, 1956. NOBLE, B.: Applied Linear Algebra, Prentice-Hall, New York, 1969. FRAZER, R. A., W. J. DUNCAN and A. R. COLLAR: Elementary Matrices, Cambridge University Press, London, 1963. WILKINSON, 3. H.: The Algebraic Eigenvalue Problem, Oxford University Press, London, 1965. FADDEVA, V. N.: Computational Methods of Linear Algebra, Dover Publications, New York, 1953. RALSTON, A., and H. S. WILF: Mathematical Methods for Digital Computers, Vol. 2, Wiley, New York, 1967. FORSYTHE, G. E., and C. B. MALER: Computer Solution of Linear Algebraic Systems, Prentice-Hall, New York, 1967. with Band Synimetric PETERS, G., and J. H. WILKINSON: "Eigenvalues of AX = A and B," Comput. J., 12, 398404, 1969.
PROBLEMS
21.
Ay =
where A and B arc symmetrical nth-order matrices and is a scalar. Suppose
B can be expressed as (see Prob. 125)
B = brb
where b is nonsingular. Reduce (a) to the form
ax =
where x
22.
let c1,
c2
by. Determine the expression for a in terms of A and b. Let x1, x2 be two nth-order column matrices or column vectors and be arbitrary scalars. If
c1x_I
+ c2x2
only when c1 =
= 0, x1 and x2 are said to be linearly independent. It follows that x1 and x2 are linearly dependent when one is a scalar multiple of the other. and Q2 arc linearly independent when Using (210) and (213), show that
23.
(a)
[3 [2
[2 0
2 7
3
(b)
to s
[3
0
24. Following the procedure outlined in Prob. 2I, determine the characteristic values and modal matrix for
+ 12Y2
l2Y1 +
62
.25.
CHARACTERISTIC-VALUE PROBLEMS
CHAP. 2
Suppose
b = p'ap
Then,
lb
t1,4
Deduce that
2(b) '(a)
k
a(b) p(a) Yk Pk
1,
2,.. ., n
1
Demonstrate for [1
21
[1 P[2
, /3,,
Q U') _. k
(a) k
in terms of q and
[3 a=[2
27.
a2 = a3 =
aa
aa2
ar =
If al # 0,
(a) (b)
a''
28.
Hint: Start with = 2,Q, and premultiply by a. A linear combination of nonnegative integral powers of a is called a
polynomial function of a and written as P(a). For example, the third order
polynomial has the form
P(a) =
c0a3
PROBLEMS
Let F(1)
F(a) =
where 0 is an nth-order null matrix. That is, a satisfies its own characteristic equation. This result is known as the Cayley-Hamilton Theorem. (a) Verify this theorem for [2 1
2
(b)
a2
f31a
/3212.
= '(a2
(c) 29.
/31a + /3213)
forn =
Establish a general expression for a ' using (225). Determine whether the following quadratic forms are positive definite.
(a) (b)
210.
F=
F = 34 +
+
+ 4x1x2 +
definite is
a11>O,a22>O
(Hint:
= 0, 211. If value of xfax1? Note that 2 0 is a characteristic value of a when a is singular. 212. Let C be a square matrix. Show that CTC is positive definite when 0 and positive sernidefinite when CI = 0. IC! (hint: Start with F = xT(CTC)x and let y = Cx. By definition, F can equal zero only when x 0 in order for the form to be positive definite.) 213. Consider the product CTaC, where a is positive definite and C is 0 and positive semisquare. Show that C1aC is positive definite when CI definite when CI = 0. Generalize this result for the multiple product,
. .
= Oforj 1,] = 1,2,.. .,n) ax = 0 has a nontrivial solution, say x1. What is. the
214. Let a be an mth-order positive defInite matrix and let C be of order m x n. Consider the product, b = CTaC
Show that b is positive definite only when the rank of C is equal to n. What can we say about b when r(C) < n? 215. Consider the quadratic form
a11
a12 a22
:
x1
a12
:
x2
x,,
a27,
64
CHARACTERISTIC-VALUE PROBLEMS
CHAP. 2
We partition a symmetrically,
(pxp)
(pXl)
NzTvTl [A11
AT (qxp)
A121 f Xt
A
(qxq)
MV ("2
where
q=
n p.
F=
Now, we take X2
> 0 for arbitrary X1, A11 must be positive definite. Since 1A111 is equal For to the product of the characteristic values of A11, it follows that Ajj must be
positive.
(a)
By taking p =
1,
2,.. .
, n,
deduce that
p=1,2,...,n
are necessary conditions for a to be positive definite. Note that it remains to show that they are also sufficient conditions.
(b)
= 0.
2-46. Refer to Prob. t25. Consider a to be symmetrical. (a) Deduce that one can always express a as the product of nonsingular lower and upper triangular matrices when a is positive definite.
(b)
j=1,2,...,n
and positive semi-definite when
j=1,2,...,n
and at least one of the diagonal elements of g is zero. Suppose we take g = hT. Then,
=
and
A
..h2i.2 A ti.11
,..i.2 Upp
Show
positive definite.
217. If a quasi-diagonal matrix, say a, is symmetrically partitioned, the submatrix A11 is also a quasi-diagonal matrix. Establish that
a=
i,j =
.
1,
.
2,. . ., N
,
PROBLEMS
65
218. Suppose we express a as the product of two quasi-triangular matrices, for example, (pxp)
(nxn)
[G11
1 [B11
B12
= [G2,
(qxp)
G22j[O
B22
where p +
= n. We take
B11
1P
8221q
Show that the diagonal submatrices of g are nonsingular for arbitrary p when a is positive definite.
Letf(x) be a function of x which is defined for the interval x1 x x2. If f(x) f(a) 0 for all values of x in the total interval x1 x x2, except x a, we say the function has an absolute minimum at x a. If f(x) f(a)> 0 for all values of x except x = a in the subinterval, x containing x = a,
we say that f(a) is a relative minimum, that is, it is a minimum with respect to all other values of f(x) for the particular subinterval. Absolute and relative maxima are defined in a similar manner. The relative maximum and minimum values of a function are called relative extrema. One should note thatf(x) may have a number of relative extreme values in the total interval x1 x x2. As an illustration, consider the function shown in Fig. 31. The relative extrema are [(a), f(h), f(c), f(d). Using the notation introduced above, we say that f(b) is a relative minimum for the interval x fib. The absolute maximum and minimum values of f occur at x = a and x = d, respectively.
f(x)
x1
x
a b
c
X2
67
In general, values of x at which the slope changes sign correspond to relative extrema. To find the relative extrema for a continuous function, we first determine the points at which the first derivative vanishes. These points are called stationary points. We then test each stationary point to see if the slope changes sign. If the second derivative is positive (negative) the stationary point is a relative minimum (maximum). If the second derivative also vanishes, we must consider higher derivatives at the stationary point in order to determine whether
the slope actually changes sign. In this case, the third derivative must also vanish for the stationary point to be a relative extremum.
Example
31
dx
and solving for x, we obtain
x2 + 4x +
x1,2 =
2
2x
+4=
2(x
2)
x2
= 2
J(x) =
The first two derivatives are
(x
a)3
+c
= 3(x
a)
a,
The stationary point, x = a, is neither a relative minimum nor a relative maximum since the third derivative is finite. We could have also established this result by considering the expression for the slope. We see from (a) that the slope is positive on both sides of x = a. The general shape of this function is shown in Fig. E3l.
66
CHAP. 3
Fig. E31
f(x)
The sufficient condition for a stationary value to be a relative extremum (relative minimum (maximum) when d2f/dx2 > 0 (< 0)) follows from a consideration of the geometry of the f(x) vs. x curve in the vicinity of the stationary point. We can also establish the criteria for a relative extremum from the Taylor series expansion of f(x). Since this approach can he readily extended to functions of more than one independent variable we will describe it in detail. Suppose we know the value of f(x) at x = a and we want f(a + Ax) where Ax is some increment in x. If the first n + 1 derivatives off(x) are continuous in the interval, a x a + Ax, we can express f(a + Ax) as
f(a + Ax)
f(a)
=
Ax +
(Ax)2
(Ax)" +
(3.1)
a, and the
where
remainder
(3-2)
a.
If f(x) is an eth-degree
polynomial, the (n + 1)th derivative vanishes for all x and the expansion will yield the exact value off(a + Ax) when n terms are retained. In all other cases, there will be some error, represented by due to truncating the series at n terms. Since depends on we can only establish bounds on The following example illustrates this point.
See Ref. 1, Article 168.
69
Example 32
We expand sin x in a Taylor series about x = 0 taking n = and noting that a = 0, we obtain sin Ax = Ax + R2
2.
If we use (a) to find sin (0.2), the upper bound on the truncation error is
0.0013.
If Ax is small with respect to unity, the first term on the right-hand side of (31) is the dominant term in the expansion. Also, the second term is more nth terms. We refer to df/dx Ax as significant than the third, fourth the first-order increment in f(x) due to the increment, Ax. Similarly, we call 4d2f/dx2(Ax)2 the second-order increment, and so on. Now, f(a) is a relative minimum when f(a + Ax) f(a) > 0 for all points in the neighborhood of
Ax e, where a, that is, for all finite values of Ax in some interval, and e are arbitrary small positive numbers. Considering Ax to be small, the first-order increment dominates and we can write
For f(a + Ax) f(a) to be positive for both positive and negative values of Ax, the first order increment must vanish, that is, df(a)/dx must vanish. Note that this is a necessary but not sufficient condition for a relative minimum, if the first-order increment vanishes, the second-order increment will dominate: f(a + Ax) f(a) =
(Ax)2
(34)
It follows from (34) that the second-order increment must be positive for > 0 to be satisfied. This requires d2f(a)/dx2 > 0. Finally, f(a + Ax) the necessary and sufficient conditions for a relative minimum at x = a are
df(a) dx
d2f(a)
dx2
a,
>
lithe first two derivatives vanish at x = the dominant term in the expansion.
(36)
Since the third-order increment depends on the sign of Ax, it must vanish for
CHAP. 3
f(a) to be a relative extremum. The sufficient conditions for this case are as
follows:
Relative Minimum
d3f d4f
dX4>
(37)
Relative Maximum
d3f
d4f
The notation used in the Taylor series expansion off(x) becomes somewhat cumbersome for more than one variable. In what follows, we introduce new notation which can be readily extended to the case of 11 variables. First, we to be the total increment in f(x) due to the increment, Z\x. define
41 = f(x +
Ax)
f(x)
(38)
and is denoted
(310)
df=-1Ax =
df(x,Ax)
The first differential off(x) is a function of two independent variables, namely, x and Ax. Iff(x) = x, then df/dx = 1 and
c/f = dx = Ax
(311)
One can use dx and Ax interchangeably; however, we will use Ax rather than dx.
Higher differentials of f(x) are defined by iteration. For example, the second
differential is given by
d2f = d(df)
Since Ax is independent of x,
(312)
=
(Ax) =
0
d2f =
(Ax)2
= d2f(x, Ax)
0.
(313)
SEC. 32.
Using differential notation, the Taylor series expansion (31) about x can be written as
(3-14)
The first differential represents the first-order increment in f(x) due to the increment, Ax. Similarly, the second differential is a measure of the secondorder increment, and so on. Then,f(x) is a stationary value when df = 0 for all permissible values of Ax. Also, the stationary point is a relative minimum (maximum) when d2f> 0 (<0) for all permissible values of Ax. The above criteria reduce to (35) when the differentials are expressed in terms of the
derivatives. Rules for forming the differential of the sum or product of functions are listed below for reference. Problems 34 through 37 illustrate their application.
f= u(x) + v(x)
df=du+dv
d2f = d(df)
d2u + d2v
15
f= u(x)v(x)
df
df = u dv + v du 2 = ud2 v + 2dudv + vd 2u
(316
3-2.
be a continuous function of n independent variables We define Af as the total increment in f due to increments in the independent variables (Ax1, Ax2
Let f(x1, (x1, x2
f(x1, x2,.. .
.
(318)
we say . , that f(x1, x2 is a relative minimum (maximum). We establish criteria for a relative extremum by expanding f in an n-dimensional Taylor series. The procedure is identical to that followed in the one-dimensional case. Actually, we just have to extend the differential notation from one to n dimensions.
72
CHAP. 3
ox2
Ax2 +
(3-19)
. ,
Ax2 where the increments are independent of (x1, x2,. The result obtained when d is applied to f is called the first
df
(320)
Higher differentials are defined by iteration. For example, the second differential has the form
d2f=d(df)=
Since
(321)
Axk
(322)
Now, we let
f(2)
j, k =
1,
2, .
. .
,n
(323)
LOXJOXkJ
Ax =
and the expressions for the first two differentials simplify to
df =
d2f =
AXTf(l)
AXTf(2) Ax
. .
.
(324)
,
The Taylor series expansion forf about (x1, x2, terms of differentials, has the form
when expressed in
Af= df +
(3-25)
j=
1,
2,. .
,n
(327)
SEC. 32.
73
called
the Euler equations for f, Note that the number of equations is equal
To summarize, the solutions of the Euler equations correspond to points at which f is stationary. The classification of a stationary point is determined by evaluated at the point. the character (definite, semidefinite, indifferent) of We are interested in the extremum problem since it is closely related to the stability problem. The extremum problem is also related to certain other problems of interest, e.g., the characteristic-value problem. In the following examples, we illustrate various special forms of f which are encountered in member
system analysis.
Example 33
f=
=
x2
y,)
df=
1
1
ek
Xk
Now,
k1
It follows that
>-&Xk
OXk
df=
Repeating leads to
j=l
d2f
f
df =
Introducing matrix notation,
>
k=t
dwfkvk +
dv1)
u=
and letting
w= =
du
[w31]
v=
{v1}
74
CHAP. 3
df =
=
d(urwv)
duTwv
+ nTdWv + uTwdv
One operates on matrix products as if they were scalars, but the order must be preserved. As an illustration, consider
f=
where
x7c
dc
and dx
Ax,
d2f =
Comparing (g) and (324), we see that
fO)
ax
=
The Euler equations are obtained by setting
ax
equal to 0:
c
The solution of (i) corresponds to a stationary value of (f). If a is positive definite, the stationary point is a relative minimum. One can visualize the problem of solving the system ax = c, where a is symmetrical from the point of view of finding the stationary XTC. value of a polynomial having the form f =
Suppose
3j7u\
\,vJ
ldu
V
UX1
:X3
1(6u
we can write
uv
vax1
df =
We apply (b) to
1.
f dv)
xx
2 AXT
(122 =
xx
Ax
AxT Ax 2 dA AxTx)
axAx=0
which we recognize as the symmetrical
problem.
(e)
SEC. 33.
LAGRANGE MULTIPLIERS
75
quotient. We have shown that the characteristic values of a are stationary values of Rayleigh's quotient. This property can he used to improve an initial estimate for a
characteristic value. For a more detailed discussion, see Ref. 6 and Prob. 311.
3-3.
LAGRANGE MULTIPLIERS
Up to this point, we have considered only the case where the function is expressed in terms of independent variables. In what follows, we discuss how one can modify the procedure to handle the case where some of the variables are not independent. This modification is conveniently effected using Lagrange multipliers. Suppose f is expressed in terms of n variables, say x1, x2 some of which are not independent. The general stationary requirement is
df= >
of
(328)
j=1
instead
establish the Euler equations, we must express df in terms of the differentials of the independent variables. Now, we suppose there are r relations between the variables, of the form
g5(x1, x2
It =
1,
2, .
,r
(329)
One can consider these relations as constraint conditions on the variables. Actually, there are only n r independent variables. We obtain r relations between the n differentials by operating on (329). Since = 0, it follows that 0. Then,
"a
j=t
k=1,2,...,r
(330)
Using (330), we can express r differentials in terms of the remaining n r differentials. Finally, we reduce (328) to a sum involving the n r indepen-
34
2
and r =
1:
f=
g(x1, x2) = 0
76
CHAP. 3
ax1
Now, we suppose ag/ax2 0. Solving (b) for dx2 (we replace dx1 by x1 is the independent variable.)
dx2= ti-----itSx1
8x2/
and substituting in (a), we obtain
/ag\
df = [PL
ax1
8x1 ox2
ox2
ax1
To determine whether a stationary point actually corresponds to a relative extremum, we must investigate the behavior of the second differential. The general form of d2f for a function of two variables (which are not necessarily independcnt) is
d2f =
2
k1
dx1 +
Of
i
= 0,
d2f
+
ax1ax.2
ox2
\0x1
ax2j Ox2
where
u=
Og Jag
The character of the stationary point is determined from the sign of the bracketed term.
An automatic procedure for handling constraint conditions involves the use of Lagrange multipliers. We first describe this procedure for the case of two variables and then generalize it for n variables and r restraints. The problem consists in determining the stationary values of f(x1, x2) subject to the constraint condition, g(x1, x2) = 0. We introduce the function H, defined by
H(x1, x2, A.) = f(x1, x2) + Ag(x1, x2)
(331)
SEC. 33.
LAGRANGE MULTIPLIERS
77
Og
A
Ox1
Ox2
(3-32)
0
OH
g(x1, x2) =
0. Then, solving the second equation in (332) for A, and substituting in the first equation, we obtain
We suppose Og/0x2
A=
0x2/ Ox2
(3-33)
and
Ox1
=
g(x1, x2) =
0 0
(3-34)
Equations (334) and (e) of the previous example are identical. We see that the Euler equations for II are the stationary conditions for f including the
effect of constraints.
Example
35
f=
g
+
x2
+ 2x1 + 7x2
=
0
We form H
f+
2g,
H=
2x1
7x2
A(x1 x2)
The stationary requirement for H treating x1, x2, and 2 as independent variables is
6x1 + 2 + 2 =
4x2
0
0
+7
x1
x2
4x2
+7
= 9/10
This procedure can be readily generalized to the case of n variables and r constraints. The problem consists of determining the stationary values Of subject to the constraints gk(xl, x2,. . , = 0, where j(x1, x2, . . , , r. There will be r Lagrange multipliers for this case, and H has k = 1, 2,
.
78
CHAP. 3
the form
H=
f + k1
+
2k9k
H(x1, x2
(335)
i = 1,2,.
k
..,n
(336)
(337)
9k = 0
= 1, 2,. . ., r
We first solve r equations in (336) for the r Lagrange multipliers, and then determine the n coordinates of the stationary points from the remaining n r equations in (336) and the r constraint equations (337). The use of Lagrange multipliers to introduce constraint conditions usually reduces the amount
of algebra.
REFERENCES
1.
Co., Reading, Mass., 1953. COURANT; R., Differential and Integral Calculus, Vol. 1, Blackie, London, 1937. COURANT, R., Differential and Integral Calculus, Vol. 2, Interscience Publishers, New York, 1936. 4. HANCOCK, H., Theory of Maxima and Minima, Dover Publications, New York, 1960. 5. APOSTOL, T. M,, Mathematical Analysis, Addison-Wesley Publishing Co., Reading, Mass., 1957. 6. CRANDALL, S. H., Engineering Analysis, McGraw-Hill, New York, 1956. 7. HILDEBRAND, F. B., Methods of Applied Mathematics, Prentice-Hall, New York,
2. 3.
1952.
PROBLEMS
31.
4x + 5 + 8x + 10 2bx + c
+ (x
a)2
4ax3
+ 4bx2 + cx + d
0, taking
32. Expand cos x in a Taylor series about x = the upper and lower bounds on R3.
n=
3.
Determine
2.
33.
Deter-
mine upper and lower bounds on R2. 34. Find df and d2f for
(a) f=x2+2x+5
(d)
PROBLEMS
35.
79
Let f =
u(x)/v(x).
Show that
(du
df =
d2f =
36. 37.
f dv)
fd2v)
Let u1, u2, u3 be functions of x and f = f(u1, u2, u3). Determine df. Suppose f = u(x)w(y) where y = y(x). Determine expressions for df and d2f. Apply to
38.
f= f' =
39.
Find the first two differentials for the following functions: + + + 5x1 4x2 + 6x1x2 + Consider f = uv, where
u=
and
Yi
u(y1, .v2)
v(y1, Y2)
y1(x1, x2)
= y2(x1, x2)
Show that
df =
d(uv)
u dv + v du
d2f =
ud2v
+ 2 du dv + vd2u
Note that the rule for forming the differential of a product is independent of whether the terms are functions of the independent variables (x1, x2) or of
dependent variables. 310. Classify the stationary points for the following functions: (a) 9x1 + 12x2 10 3xl +
(b)
f f f
f
311.
x is arbitrary. Since a is symmetrical, its characteristic vectors are linearly independent and we can express x as
x=
where Q3
(a)
n)
= j=j-
80
CHAP. 3
<<
Suppose x differs only slightly from Qk. Then, ICjI Specialize (a) for this case. Hint: Factor out 2k and Use (b) to obtain an improved estimate for A.
for j
k.
a=[i
x
[3
{1, 3}
2=1
x={1,2}
312. Using Lagrange multipliers, determine the stationary values for the following constrained functions:
(a) g
(b)
x2
g1 =
313. Consider the problem of finding the stationary values of f = = xrarx subject to the constraint condition, = 1. Using (336) we write
g2=x1x2+2x3+2=0
H =f+ Ag =
x1 + x2 + X3
2(XTX
1)
(a)
Show that the equations defining the stationary points off are
ax=Ax
(b)
xTx=1
Relate this problem to the characteristic value problem for a symmetrical matrix. 314. Supposef = a. Show that and g = I xTax = 0 where aT the Euler equations for H have the form
xTax=1
We see that the Lagrange multipliers are the reciprocals of the characteristic values of a. How are the multipliers related to the stationary values of f?
A curve is defined as the locus of points whose position vector* is a function of a single parameter. We take an orthogonal cartesian reference frame having and X3 (see Fig. 41). Let F he the position Vector to a point directions X1,
x3
i3
X3(y)
x2
X2(y)
x1
81
82
CHAP. 4
=
an alternate representation is
Since F =
ii
(j =
1,2, 3)
(42)
Both forms are called the parametric representation of a space curve. Example
(1)
4i
let a =
Consider a circle in the X1-X2 plane (Fig. E41A). We take y as the polar angle and The coordinates are
and
x1 = x2 = F =
a
a
cos y sin y
+ asiny'12
=
X3 =
cos y
(43)
x2 = bsiny
CY
where a, b, care constants. The projection on the X, -X2 plane is an ellipse having semiaxes a and b. The position vector for this curve has the form
F
= a cos
42.
ARC LENGTH
and The cartesian coordinates are length of the chord from P to Q is given by
y+
(j
1,
2, 3) and the
As Exy * 0,
In the limit,
Noting that
=
dx1 =
dy
we can express ds as
ds
+
dy
(4-4)
Fig. E41A
Fig. E41B
Q(+Ay)
I LISI
P(y)
Fig. 42. Differentia' segment of a curve.
83
84
CHAP. 4
dx
dx
21/2
dy
(45)
We have defined ds such that s increases with increasing y. It is customary to call the sense of increasing s the positive sense of the curve. To simplify the expressions, we let
=
+
Then, the previous equations reduce to
dx
1/2
(4-6)
ds =
dy
(47)
One can visualize as a scale factor which converts dy into ds. Note that +1. x > 0. Also, if we take y = s, then
Example
42
Consider the curve defined by (43). Using (46), the scale factor is
[a2 sin2 y +
We suppose that b
b2
cos2 y + c2]'12
a. One can always orient the axes such that this condition is satisfied.
as
Then, we express
where
= (b2 + c2)"2 [1
k2 = b2 b2 +
The arc length is given by
s
2
k2
sin2 y]"2
c2
dy =
(b2
[1 k2 sin2 yJ112 dv
The integral for s is called an elliptic integral of the second kind and denoted by E(k, y).
Then,
s
(b2 +
c2)'12
E(k, y)
a,
Tables for E(k, y) as a function of k and y are contained in Ref. 3. When b = is called a circular helix and the relations reduce to
the curve
(a2
const.
S = ny
SEC. 43.
43.
85
We consider again the neighboring points, P(y) and Q(y + shown in Figure 43. The corresponding position vectors are P(y + ky), and
approaches the tangent to the curve at P. Then, the unit tangent vector at P is given by*
As L\y -+ 0,
t=
Jim
PQ d1 --=ds
(48)
dP
dy
dF
ds
dy ds
(49)
dy
Since > 0, 1 always points in the positive direction of the curve, that is, in the direction of increasing s (or y). It follows that dP/dy is also a tangent vector and
dy
/df dP\"2
\dy dy
Equation (410) reduces to (46) when coordinates.
+s
(410)
is expressed in
cartesian
Q(y+6.y)
r(y)
Fig. 43. Unit tangent vector at P(y).
86
CHAP. 4
Example 43
We determine the tangent vector for the curve defined by (43). The position vector is
F
= a cos
dy
a sin
+ h cos Y12 +
dy
It follows from (a) that di/dy is orthogonal to f. The unit vector pointing in the direction of di/dy is called the principal normal vector and is usually denoted by ii.
H=
ldt
dy
where
d (1
The binormal vector, h, is defined by
(4..-lt)
dF
b='?xh
We see that b is also a unit vector and the three vectors.
,
(4-12)
b comprise a right-
handed mutually orthogonal system of unit vectors at a point on the curve. Note that the vectors are uniquely defined once y) is specified. The frame
associated with b_and ii is called the moving trihedron and the planes determined by (1, ), (ii, b) and are referred to as the osculating normal, and
rectifying planes (see Fig. 44).
* See
Ref. 4, Chap.
1.
SEC. 44.
87
Normal plane
Rectifying
plane
Example 44
We determine fi and b for the circular helix. We have already found that
a [a2 +
and
sin VT1
+ a cos
+ c13]
di
Then,
a [cos ytj + Sm
i dt
fl
dt dy
dy
C05
Sm
The principal normal vector is parallel to the plane and points in the inward radial direction. It follows that the rectifying plane is orthogonal to the X1-X2 plane. We can determine b using the expansion for the vector product.
a
This reduces to
asiny acosy
C.
a
sin
cos
a
3
88
CHAP. 4
Fig, E44
45.
The derivative of the tangent vector with respect to arc length is called the curvature vector, K.
K
dt
c/s
d2F
i/s2
K
Using (411), we can write
ic/i
c/s2
(413)
ds
Ku
(414)
0. The Note that K points in the same direction as Ft since we have taken K curvature has the dimension L1 and is a measure of the variation of the tangent vector with arc length. We let R be the reciprocal of the curvature:
R=
K1
iS)
In the case of a plane curve, R is the radius of the circle passing through three consecutive points* on the curve, and K = JdO/dsj where 6 is the angle between I and To show this, we express I in terms of 0 and then differentiate with respect to s. From Fig. 45, we have
cos
* See
+ Sifl 012
SEC. 45.
Then
K
and
[sin
dO
1
+ cos 617]
dO a
ds
dO/ds
R
[ sin
+ cos 612]
In the case of a space curve, the tangents at two consecutive points, say P and Q, are in the osculating plane at F, that is, the plane determined by and at. P. We can interpret R as the radius of the osculating circle at P. It should be noted
that the osculating plane will generally vary along the curve.
x2
\
R
+
R
t
i2
it
The binormal vector is normal to both and and therefore is normal to the osculating plane. A measure of the variation of the osculating plane is given by db/ds. Since his a unit vector, db/ds is orthogonal to h. To determine whether db/ds involves we differentiate the orthogonality condition I b 0, with
respect to s.
-
db
dl
ds
ds
But dl/ds K and b ii = 0. Then, db/ds is also orthogonal to I and involves only . We express db/ds as
db
= tn
(416)
90
CHAP. 4
xn
Differentiating with respect to s, we have
db = di
xn+t
diii
This reduces to
=t
x
db
ii =
0.
l-dfl
dy
(417)
Note that a can be positive or negative whereas K is always positive, according to our definition. The torsion is zero for a plane curve since the osculating plane coincides with the plane of the curve and b is constant.
Example
45
[a
cos
sin vij +
sin
b=
where
a = (a2 +
Then,
c2)112
K=- adv
and
1
a
a
a +c
c c
2
dy
2const
We have developed expressions for the rate of change of the tangent and binormal vectors. To complete the discussion, we consider the rate of change of the principal normal vector with respect to arc length. Since fi is a unit vector, d/ds is orthogonal to . From (417),
b ds
SEC. 46.
91
orthogonality relation, I n = 0.
ds
ds (b)
= I(t + tb
us
(418)
The differentiation formulas for 1, ii, and b are called the Frenet equations.
46.
We
di
thu
1
= = tangent vector
exdy
ldi
a=
i-- i
di
= I x binormal vector di ds
dy dy
Di:fferenriation Formulas
Equations)
dl
lull
+tb
(420)
K=
=
1 di curvature a dy 1d
a
= torsion
dy
We use the orthogonal unit vectors (I, , b) to define the local reference frame for a member element. This is discussed in the following sections. The Frenet
92
CHAP. 4
member element.
LOCAL REFERENCE FRAME FOR A MEMBER ELEMENT
47.
The reference frame associated with , and b at a point, say P, on a curve is uniquely defined once the curve is specified, that is, it is a property of the
curve. We refer to this frame as the natural frame at P. The components b) are actually the direction cosines for the natural of the unit vectors
frame with respect to the basic cartesian frame which is defined by the orthogonal unit vectors (1k, 12, 13). We write the relations between the unit vectors as ft
n
12
133
11
t22
32
e33
12
(421)
One can express* the direction cosines in terms of derivatives of the cartesian
coordinates (x1, x2, x3) by expanding (419). Since (1, b) are mutually or13) the direction cosines are related by thogonal unit vectors (as well as
1jm6m
j,
1,
2, 3
(422)
and we see that is an orthogonal matrix.f The results presented above arc applicable to an arbitrary continuous curve. Now, we consider the curve to be the reference axis for a member clement and take the positive tangent direction and two orthogonal directions in the normal plane as the directions for the local member frame. We denote the directions
of the local frame by (Y1, Y2, 1'3) and the corresponding unit vectors by (t1,
2,
= 1) We will always take the positive tangent direction as the Y1 direction x and we work only with right handed systems t3). This notation is shown in Fig. 4-6. When the centroid of the normal cross-section coincides with the origin of the local frame (point P in Fig. 46) at every point, the reference axis is called the centroidal axis for the member. It is convenient, in this case, to take Y2, Y3 as the principal inertia directions for the cross section. In general, we can specify the orientation of the local frame with respect to the natural frame in terms of the angle between the principal normal direction and the I'2 direction. The unit vectors defining the local and natural frames
* See
Prob. 4-5.
SEC. 47
93
are
related by tl
t2
COS 4)11
+ sin 4)b
(424)
bn + cos 4th
Combining (421) and (424) and denoting the product of the two direction cosine matrices by the relation between the unit vectors for the local and basic
t=
where
(425)
[
21cos4)+31sin4)
12
j3
22cos4)+ 32sin4)
23cosb+ 33sinqS
23 sin 41+ 33cos41
[21sin4)+31cos4)
22si+C32cos4)
Note that the elements of fi are the direction cosines for the local frame with respect to the basic frame.
fJjk =
Xk)
(426)
'. We will utilize (425) in the next Since both frames are orthogonal, J1 chapter to establish the transformation law for the components of a vector.
x3
Normat
Y1
Fig. 46. Definition of local reference frame for the normal cross section.
94
CHAP. 4
Example 46
We determine
for the circular helix. The natural frame is related to the basic frame.by
I
slay
a
a.
cosy
a
a
0
a
=
b
cos
sin y
c
12
{Ik}
C. slay
a
--cosy a
Using (4-.25)
a a
a cos y a
sinycosb cosysin4
a
a
+cosysm4 + sinycos4
C.
sin y sin
cos
y cos
I. a
a
a
48.
We take as curvilinear coordinates (yi, Yz' y3) for a point, say Q, the parameter
of the reference axis and the coordinates (Y2, of Q with respect to the orthogonal directions (Y2, 1'3) in the normal cross section (see Fig. 47). Let F(y1) the R(y1, Y2' Y3) be the position vector for Q(Yl, y3) and position vector for the reference axis. They are related by = r + Y2t2 + y3t3
where
COS
+ Sifl (
t3 =
+ cos4b
(427)
y3
Y2
SEC. 48.
95
The curve through point Q corresponding to increasing Yj with Y2 and y3 held constant is called the parametric curve (or line) for yj. In general, there are three parametric curves through a point. We define as the unit tangent vector for the parametric curve through Q. By definition,
Ui
13R
(428) aIR
=
The differential arc length along the
aIR
curve is related to
by
(429)
=
(or
This notation is illustrated in Fig. 48. One can consider the vectors
to define a local reference frame at Q.
x3
y2t2 +y3t3
x2
dy1
=
=
t2
t3
dt2
Y2
dy1
+ Y3
dy1
aR aR
96
CHAP. 4
We see that
t2
g2
3=t3
It remains to determine 1 and g1. Now,
dy1
g3=1
(430)
=
1db '\dv1
Also,
d12
(d
\dy1
dy1
+ dy1 + bj
db
dy1
11 dy1
dt3
(dii
\dy1
+ b)+ dy1j
dq5\
1db
\dy1
dy1
dyj
I
and finally,
= cc(1
d4)'\.
Ky'2)!1
d4)\ _)(Y2t3
y3t2)
(431)
Y2 COS 4)
J73 sin 4)
We see from Fig. 49 that y'2 is the coordinate of the point with respect to the
\y3
Fig. 49.
of y.
Since 13R/ay1 (and therefore ii1) involve the reference frame defined and by iii, u2, will not be orthogonal. However, we can reduce it to an orthogonal
REFERENCES
97
system
by taking
dy
=
cer dy
150
(432)
which requires
=
When (432) is satisfied,
(433)
aR
and
Ky'2)t1
= =
(434)
cx(l
In this case, the local frame at Q coincides with the frame at the centroid. One
should note that this simplification is practical only when ccc can be readily
integrated.
Example
47
a=
(a2
C
+ c2)112
Then,
at =
a
tS
varies linearly with y (or arc length). The parameter g1 follows from (434).
hi =
ds1
= a(1
Ky2)
a
cc-
x(l
'\
REFERENCES
1.
2.
3.
THOMAS, G. B., JR.: Analytical Geometry and Calculus, Addison-Wesley Publishing Co., Inc., Reading, Mass., 1953. HAY, U. 13.: Vector and Tensor Analysis, Dover Publications, New York, 1953. JA}INKE, E., and F. EMDE: Tables of Functions, Dover Publications, New York, 1943.
4.
CHAP. 4
PROBLEMS
41.
(a)
(b) (c) (d)
Il, b, ; K, x for the following curves: x2 = 3 sin y cos y x3 = 5y x1 3 cos y x2 = 6 sin y x3 = 5y + p313 + = x1 = cos y x2 = sin y
Determine
3
x1 =
x2=J(x1)
x3=0
Equation (a) corresponds to taking x1 as the parameter for the curve. (a) Determine the expressions for 7, , b, and K corresponding to this representation. Note that Let y and + f(x1)12 +
_Lf'
(b)
.f"etc
4a
2 x1)
=
where
(c)
a and b are constants. This is the equation for a parabola symmetrical about x1 = b/2. Let 9 be the angle between and
cos0 = I
Deduce that
Specialize
(d)
11.
= sec 0. Express t, h, !, and K in terms of 0. for the case where 02 is negligible with respect to unity. This approximation leads to
sin 0 cos 0
tan 0
1
43.
A curve is said to be shallow when 02 << 1. Show that (see (420)) Let K = l/R and t
di
dh d
= 12
dy.
PROBLEMS
44.
99
x1 =
or
cos y
2 x1
2 x2
x2 =
sin y
Determine 1, n for both parametric representations. Take x1 as the parameter for (b). Does y have any geometrical significance? 45. Show that
dx,.
72
1k
IA
C2k
[3 (d [,.=,. \ dyj J
13
dy
dv2
=
42
\ dy
121
dy dyj j
311 32
33j
46.
42
Let
43]
Then,
1
[4k]
3
Ii' iT
ii
Determine D for Prob. 4la. Determine JI for Prob. 4lb. 4-9. Specialize for the case where the reference axis is in the X1 X2 plane. Note that b = 3313 where 431 = 1. When the reference axis is a plane curve and = 0, we call the member a "planar" member. as 410. We express the differentiation formulas for
48.
dt = at ds
(a) (b)
Show that a is, in general, skewsymmetric for an orthogonal system of unit vectors, ie., c5Jk. Determine a. Suppose the reference axis is a plane curve but b 0. The member is not planar in this case. Determine a.
ROTATION TRANSFORMATION
Suppose we know the scalar components of a vector with respect to a reference frame and we want to determine the components of the vector corresponding to a second reference frame. We can visualize the determination of the second set of components from the point of view of applying a transformation to the column matrix of initial components. We refer to this transformation as a rotation transformation. Also, we call the matrix which defines the transformation a rotation matrix. (j = 1, 2, 3 and n = 1, 2) be the directions and corresponding unit Let vectors for reference frame n. (See Fig. 51.) We will generally use a superscript to indicate the reference frame for directions, unit vectors, and scalar
SEC. 51.
ROTATION TRANSFORMATION
a=
(a2)Ti2
To proceed further, we must relate the two reference frames. We write, the relations between the unit vectors as
i2 = Ili' where is the scalar component of with respect to The transformation matrix, is nonsingular when the unit vectors are linearly independent. Substituting for and equating the coefficients of i' leads to
a' =
a2
Finally,
we let
R'2 R21 =
form
(53)
With this notation, the relations between the component matrices take the
a2 = a1 = =
R21a2
formation. For example, R'2 is the rotation transformation matrix corresponding to a change from frame 1 to frame 2. We see that the transformation matrix for the scalar components of a vector is the inverse transpose of the transformation matrix governing the unit vectors for the reference frames.
Example 5i
We consider the two-dimensional case shown in Fig. E5-1. The relations between the unit vectors are = cos + sin = sin + cos
cos6 sinO
102
CHAP. 5
Then,
R21
sin4 Lsinfl
1
cos4
Lsjn 6
cos U
R'2 = (fV)
=
+
sin
o sin
0 cos
and
Jafl
cos
?J L sin 0
cos oJ
ir
Fig. E51
12
The result obtained in the preceding example can be readily extended to the
orthogonal reference frames. When both frames are orthogonal, the change in reference frames can be visualized as a rigid body rotation of one frame into the other, f3jk is the direction cosine for with respect to and the rotation transformation matrix is an orthogonal matrix:
case of two
1 R'2 LI'jki
A' k
(55)
In Sec. 47, we defined the orientation of the local frame (1k, 13, at a point on the reference axis of a member element with respect to the natural frame (1, , b) at the point. This frame, in turn, was defined with respect to a fixed cartesian frame 12, 13). In order to distinguish between the three frames, we use superscripts p and p' for the local and natural frames at p and a superscript 1 for the basic cartesian frame: = = {Z, =
t2,
ii, b}
(56)
SEC. 52.
103
With this notation, the relations between the unit vectors and the various
= R141
=
1.
From(421),
(57)
1
2.
From (424),
0 0
=0
0
3.
defined by (425).
52.
The equilibrium analysis of a member element involves the determination of the internal force and moment vectors at a cross section due to external forces and moments acting on the member. We shall refer to both forces and moments as "forces." Also, we speak of the force and moment at a point, say P, as the "force system" at P. The relationship between the external force
system at P and the statically equivalent internal force system at Q hasa simple form when vector notation is used. Consider a force F and moment M acting at P shown in Fig. 52. The statically equivalent force and moment at Q are
Feqrnv.
Mequiv = M +
XF
--
(5-8)
One can visualize (58) as a force transformation in which the force system
Fcquiv.
linear if
is constant, that is, if the geometry of the element does not change
appreciably when the external loads are applied. We will write (58) in matrix form and treat force transformations as matrix transformations.
104
CHAP. 5
We develop first the matrix transformation associated with the moment of a force about a point. Let be a force vector acting at point P and MQ the moment vector at point Q corresponding to We will always indicate the point of application of a force or moment vector with a subscript. The relation
between MQ and
is
= QP x (59) We work with an orthogonal reference frame (frame 1) shown in Fig. 53 and write the component expansions as
v' PL.. ri
1
._
'1
(jl)TM1
(510)
(1
(42
1\
XQ2)
(1
(1.
0
Note that matrix. One can interpret it as a forceis a into translation transformation matrix. The force at P is transformed by
F1 P3
'Ik
i-i
,1
M1
43
I'
/
Xp3
/ xP1
/ 'i
SEC. 52.
a
105
moment at Q. Note that the order of the subscripts for the translation transcorresponds to the order of the translation (from P formation matrix, and must be referred to the same frame, that is, the superto Q). Also, scripts must be equal. Up to this point, we have considered only one orthogonal reference frame. In general, there will be a local orthogonal reference frame associated with each point on the axis of the member, and these frames will coincide only when the member is prismatic. To handle the general case we must introduce rotation transformations which transform the components of F and M from the local frames to the basic frame (frame 1) and vice versa. We use a superscript p to indicate the local frame at point P and the rotation matrix corresponding to a transformation from the local frame at P to frame 1 is denoted by R'1. With this notation, =
=
and the general expression for takes the form
(512)
(513)
We consider next the total force transformation. The statically equivalent force and moment at Q associated with a force and moment at P are given by
MQ =
+ QP x
(a)
When all the vectors are referred to a common frame, say frame 1, the matrix transformation is
'Q)
L"PQ
L3J
(b)
We let
(514)
is called the force system at Q referred to frame 1. Using The 6 x 1 matrix this notation, (b) simplifies to d1
When the force systems are referred to local frames, we mast first transform them to a common frame and then apply (5--15). Utilizing the general
matrix,
I
(516)
and applying.
(a)
106
CHAP. 5
we
obtain
(517)
is applied Equation (517) states that when the matrix transformation we obtain its statical equivalent at Q. Actually, we could leave off the to appears subscripts and superscripts on when we write (517). However,
alone, we must include them. Note that the force transformation generally involves both translation and rotation. The order of the subscripts corresponds to the direction of the translation, e.g., from P to Q. Similarly the order of the
superscripts defines the direction of the rotation or change in reference frames, e.g., from frame p to frame q. In general, the geometry of a member element is defined with respect to a basic we must determine reference frame which we take as frame 1. To evaluate from the geometrical relations for the member. We have and Rn', already discussed how one determines R1 in Secs. 47 and 5f.
When the member is planar* and the geometry is fairly simple (such as a straight or circular member), we can take frame I parallel to one of the local frames. This eliminates one rotation transformation. For example, suppose we reduces to take frame I parallel to frame p. Then, R1 = 1 and
[R
(S18)
reduces
I
to
(5-19)
(520)
(521)
S2,...,
matrices.
Q, where
tion matrix,
is carried out in the order P S1, S1 are intermediate points, the transformais equal to the product of the intermediate transformation
S2,
.
.
, S,,
(522)
* If the reference axis isa plane curve and the local frame coincides with the natural frame we say the member is planar.
0)
SEC. 52.
107
, SN
Example
52
We consider the plane circular member shown in Fig. ES2. We take frame 1 parallel to frame p. Then,
= {a sin 0, a(1
0
cos 0), O}
a(1cos6)
a
a(1 cos 0)
0 a sin 0
0
Sin 6
cos6
=
= sin 0
0
sinO cos 0
0
9
I
where
0 0
=0
a(1 cos 0)
asin0
P
p
2
t7
As an illustration of the case where the geometry is defined with respect to a basic for a circular helix. The general
108
CHAP. 5
expansion for
= =
where
cos y sin y
= cy
(j = 1,2,3) are the cartesian coordinates with respect to the basic frame (frame 1). Let yp and y corresponding to points P and Q. The coordinate matrices for P and Q are
Cyp)
YQ)
C(yp
y0)
a(sin
sin yQ)
a(cos Yp
To simplify the algebra, we suppose the local frame coincides with the natural frame at every point along the reference axis, that is, we take = 0. Using the results of Sec. 47, the rotation matrices reduce to
a
a
C
C.
a
cc
C
cc
where cc2
we obtain
(a\2
II
/c\2
I
a sin ij
cc
I
i (1 cos
cc
ac
c
cc
/a\2
/c'\2
\ccJ
= yp
=
y0. Also,
a2c 3-(2
ac
a
(IC
. I
sinq)
cosn)(a5
c2)]
ac2
a
c2)
a
a2
a
ci2c
Note that we can specialize the above general results for the case of a plane circular member (Example 5-2) by taking c = 0 and = 0.
53.
Let P and Q be two points on a rigid body. Suppose that the body experiences a translation and a rotation. We define Up and Ui,, as the translation and rotation* vectors for point P. The corresponding vectors for point Q are given by
UQ = Up + (Up X
(523)
= Wp
is
PQ x
UQ = Up + QP X
(525)
[13
L"l 3J
=
0
XPQ1
i
consider next the case where the local frames at P and Q do not coincide. The general relation between the displacements has the form
13
(527)
[o
One can showt that alternate forms of (527) are
j
T
d14
=
The units
are
(yqp)T
radians.
110
CHAP. 5
HALL, A. S., and R. W. WOODS-LEAD: Frame Analysis, 2d ed., Wiley, New York, 1967.
2. 3. 4.
5.
MORLCE, P. B.: Linear Structural Analysis, Ronald Press. New York, 1959. PESTEL, E., and LECKIE, F.: Matrix Methods in Elastomechanics, McGraw-Hill, New York, 1963. Livasr.Ey, R. K.: Matrix Methods of Structural Analysis, Pergamon Press, London,
1964.
MARTLS.t, H. C.: Introduction to Matrix Methods of Structural Analysis, McGraw-Hill, New York, 1966.
PROBLEMS
51.
a1 =
52. The orientation of two orthogonal frames is specified by the direction cosine table listed below.
1/2 1/2
(a)
1/2 1/2
\/2/2
Determine R12. Verify that (R12)T = (R12y1
PROBLEMS
(b) If a1 = {10, 5, 10}, find a2. (c) If a2 {5, 10, 10), find a'. 53. Consider two points, P and Q, having coordinates (6, 3,2) and (5, 1,4)
with respect to frame 1. The direction cosine tables for the local reference frames are listed below.
1
1/2
1/2
1/2 1/2
1/2
1/2
1/2
1/2
12/2
(a) (b)
(c)
Determine Suppose
5--4.
{l00, 50, 100, 20, 40, +60). Calculate Consider the planar member consisting of a circular segment and a
Prob,. 54
S
straight segment shown in the sketch below. Point P is at the center of the circle.
t'i
Q
Ic
(a) (b)
(c)
55.
Determine by transforming directly from P to Q. Also find Determine by transforming from P to S and then from S to Q. Find corresponding to = {0, 0, 1, 0, 0, 0}. Consider the circular helix,
=
(a)
cos
+ 2 sin
+
Take
13.
Suppose 4(y) =
Suppose
0.
(b)
y.
Determine Determine
= ir/2, YQ =
112
CHAP. 5
56.
corresponding to
P
= {1/2,
Q_
I _T___
57.
Verify that (527) and (528) are equivalent forms. Note that
1
rI
IA
ii
58.
3 3J L" 1 Consider the plane member shown. The reference axis is defined by
niT
I I
ort
x2
= f(xj).
Prob. 58
x1
(a)
(b)
Determine Note that the local frame at P coincides with the basic frame whereas the local frame at Q coincides with the natural frame at Q. Specialize part (a) for the case where
4a
2 (x1b Xj)
coordinate of point Q
is
XCI)
Cl)
GENERAL
A system of bars* connected at their ends by frictionless hinges to joints and subjected only to forces applied at the joint centers is called an ideal truss.t The bars arc assumed to be weightless and so assembled that the line connecting the joint centers at the ends of each bar coincides with the centroidal
axis. Since the bars are weightless and the hinges are frictionless, it follows that
each bar is in a state of direct stress. There is only one force unknown associated with each bar, namely, the magnitude of the axial force; the direction of the force coincides with the line connecting the joint If the bars lie in one plane, the system is called a plane or two-dimensional truss. There are two displacement components associated with each joint of a plane truss. Similarly, a general system is called a space or three-dimensional truss, and there are three displacement components associated with each joint. We suppose there are in bars (members) and j joints. We define i as
= =
2
3
Using this notation, there are if displacement quantities associated with the] joints. In general, some of the joint-displacement components are prescribed. Let r be the number of prescribed displacement components (displacement restraints) and nd the total number of unknown joint displacements. It follows that = if r (62)
Corresponding to each joint displacement restraint is an unknown joint force
A prismatic member is conventionally referred to as a bar in truss analysis.
See Ref. 1.
115
116
CHAP. 6
n1m+r
Finally, the total number of unknowns, n, for an ideal truss is
(63)
flflj+flj(j+tfl
(64)
The equilibrium equations for the bars have been used to establish the fact that the force in each bar has the direction of the line connecting the joint
centers at the ends of the bar. There remains the equilibrium equations for the joints. Since each joint is subjected to a concurrent force system, there are iJ scalar force-equilibrium equations relating the bar forces, external joint loads, and direction cosines for the lines connecting the joint centers in the deformed state. In order to solve the problem, that is, to determine the bar forces, reactions, and joint displacements, m additional independent equations are required. These additional equations are referred to as the bar forcejoint displacement relations and are obtained by combining the bar forcebar elongation relation and bar elongationjoint displacement relation for each of the in bars. In this chapter, we first derive the elongationjoint displacement relation for a single bar and then express the complete set of in relations as a single matrix equation. Thisprocedure is repeated for the bar force-elongation relations and the joint force-equilibrium equations. We then describe a procedure for introducing the joint-displacement restraints and summarize the governing equations. Finally, we briefly discuss the solvability of the governing equations for the linear case. In this case, the question of initial instability is directly related to the solvability.
We prefer to proceed from the truss to the general system since the basic formulation techniques for the ideal truss can be more readily described. To
adequately describe the formulation for a general system requires introducing a considerable amount of notation which tends to overpower the reader.
62.
We number the joints consecutively from 1 through j. It is convenient to refer the coordinates of a joint, the joint-displacement components, and the external joint load components to a common right-handed cartesian reference frame. Let (j = 1, 2, 3) be the axes and corresponding orthogonal unit
SEC. 62.
vectors for the basic frame. The initial coordinates, displacement components,
and components of the resultant external force for joint k are denoted by
(j =
1,
rk =
j= Uk = Ukl
Pk
1
(65)
The coordinates and position vector for joint k in the deformed state are
1k
11k
(66)
+ Uk
14313
Deformed position
of joint k
//
11k2
//
x2
flkl
x1
We number the bars from 1 through m and consider bar n to be connected to joints k and s. The centroidal axis of bar n coincides with the line connecting joints k and s. From Fig. 62 the initial length of bar n, denoted by = is equal to the magnitude of the vector
(67)
116
CHAP. 6
= Before the orientation of the bar can be specified, a positive sense or direction must be selected. We take the positive sense for bar n to be from joint k
13
xk)T(xs xk)
xkf)
(68)
x.,3
/ / XkI /
Xg3
//
12
Xk2
to joint s and define as the direction cosine for the positive sense of bar n direction: in the undeformed state with respect to the
(Ar
XkJ)
(69)
=L
Xk)
(610)
1, due to the orthogonality of the reference frame. Finally, Note that we let be the unit vector associated with the positive direction of bar n in
the undeformed state. By definition,
-
Ar =
12
The deformed position of bar n is shown in Fig. 63. The length and direction cosines for bar n are equal to the magnitude and direction cosines for the vector, the unit vector = Pk. Let L12 + e12 be the deformed length,
SEC. 62.
x3
Ap
Joints
x2
t?&2
x1
associated with the positive direction in the deformed state, and the corresponding direction cosine matrix. These quantities are defined by
+
eh)
1
(612)
(613)
(614)
Uk)
(1 +
2.
Uk) +
uk)T(US
ilk)
(615)
expands to
[cx,,j
I +-;:-
UkJ)
(616)
120
CHAP. 6 (617)
[cm
uk)T]
By definition, is the change in length of bar n. Then, e,,/L7, is the extensional strain which is considerably less than unity for most engineering materials. For example, the strain is only for steel at a stress level of 3 x ksi. The relations simplify if we introduce the assumption of small strain,
<< I
Uk) +
(ii,, uk)T(u$
Uk)
(619)
Uk)
(620)
To simplify the expression for further, we need to interpret the quadratic terms. Using (620), we can write (619) as
=
This
(u,
Uk)
form shows that the second-order terms arc related to the change in
orientation of the bar. If the initial geometry is such that the bar cannot experience a significant change in orientation, then we can neglect the nonlinear terms. We use the term linear geometry for this case. The linearized relations are
Uk)
(621)
We discuss this reduction in greater detail in Chapter 8. Since we are concerned in this chapter with the formulation of the governing equations, we will retain the nonlinear rotation terms. However, we will assume small strain, i.e., we
work with (619), (620).
63.
We have derived expressions for the.direction cosines and elongation of a bar in terms of the initial coordinates and displacement components of the joints at the ends of the bar. By considering the truss as a system or network, the geometric relations for all the bars can be expressed as a single matrix equation. The relations for bar n, which is connected to joints s and k (positive direction
121
= (x. = = =
=
Xk)
(x., xk)
Uk)
+ +
(ii, uk)T
Uk)
is more convenient for generalization of the geometric relations. Let n, n_ denote the joint numbers for the joints at the positive and negative ends of and k member n. The geometric relations take the form (we replace s by
bynJn(a)):
=
=
x,,.)
=
= ci,,
(622)
(u,,,
11)T
ci,,
u,,)
To proceed further, we must relate the bars and joints of the system, that is, we must specify the connectivity of the truss. The connectivity can be defined by a table having m rows and three columns. In the first column, we list the bar numbers in ascending order, and in the other two columns the corresponding numbers, and n, of the joints at the positive and negative ends of the members. This table is referred to as the branch-node incidence table in network theory.* For structural systems, a branch corresponds to a member and a node to a joint, and we shall refer to this table as the member-joint incidence table or simply as the connectivity table. It should be noted that the connectivity depends only on the numbering of the bars and joints, that is, it is independent of the initial geometry and distortion of the system.
Example
61
As an illustration, consider the two-dimensional truss shown. The positive directions of the bars are indicated by arrowheads and the bar numbers are encircled. The connetivity
See Ref. 8.
122
CHAP. 6
table (we list it horizontally to save space) for this numbering scheme takes the following
form: Bar,n
1
2 2
3
4
5
6 2
5
9
2 6
10
4 2
11 5
3
+Joint(n)
Joint(n..)
4
5
Fig. E61
0
6
With the connectivity table, the evaluation of the initial length and direction cosines can be easily automated. The initial data consists of the j coordinate matrices, x1, x2 To compute and a,,, we first determine n and n_ from the connectivity table and then use the first two equations of(622). For example, for bar 8, 8 5, and 1, 8..
xs
x1 x5
(x1
x)T(x * x5)
x5)T
e=
{e1,
C2,...,
= {u1,u2
623
e = u
(624)
where d is of order m x ij. The elements in the uth row of d involve only Then, partitioning d into submatrices, of order 1 x 1, the elements of where k = 1, 2 in and = 1, 2 it follows that the only nonvanj, ishing submatrices for row n are the two subinatrices whose column number
corresponds to the joint number at the positive or negative end of member it,
namely,
n and n.:
=
+777
(625)
t'
=
Example
62
when
n orn
The .& matrix can be readily established by using the connectivity table. For row n, one puts +y,, at column at column n_, and null matrices at the other locations. The general form of the d matrix for the truss treated in Example 61 is listed below. We have also listed the elongations and joint displacement matrices to emphasize the significance of the rows and partitioned columns of .&.
Uj
e1
U2
U3
U4
U5
U6
Ii
0
12
0
0
0 0
13
0
0
e2
e3
0 0 0
13
0
14
e4
0
15
0
0 0
14 0
e5
e6
0
0
is
0
0
0
Ys
0 0
e5 e9
e10
is
0
0
19
0 0
0
0
0
19
0
0
1io
0
ho
0
0 0
e11
The d matrix depends on both the geometry and the topology. It is of interest to express d in a form where these two effects are segregated. The form of (625) suggests that we list the y's in a quasi-diagonal matrix,
71
72
(626)
124
CHAP. 6
and define C as
C = [Ck(i
1,2
in
e= 1,2,...,j
Cr,, -
*= +
=0
Then,
= Ij
n or n...
(627)
d = yC
(628)
The network terminology* for C is augmented branch-node incidence matrix. We shall refer to it simply as the connectivity matrix.
Example 63
The connectivity matrix for Example 61 is listed below. The unit matrices are of order 2 since the system is two-dimensional.
Joint Numbers
1
12
+12
12
+12
12
Bar Numbers
+12
+12 +12 +12
12 12 12
+12
12
+12
One can consider row n of C to define the two joints associated with bar n. It follows that column k of C defines the bars associated with joint k. This association is usually
See Prob. 66.
SEC. 64.
125
referred
the positive end of the bar. Similarly, a bar is positive incident on a joint when its positive
end is at the joint. For example, we see that joints 1 and 4 are incident on bar 5 and
bars 3, 4, 6, 8, and 11 are incident on joint 5. We will use this property of the connectivity matrix later to generalize the joint force-equilibrium equations.
C
64.
Fig.
A = cross sectional area F axial force, positive when tension = initial elongation, i.e., elongation not associated
with stress
*
126
CHAP. 6
Since the stress and strain are constant throughout the bar,
F=
e = e0 =
Lg
Le0
(629)
We convert the a-c relation for the material to the force-elongation relation
for the bar by applying (629).
C0
We consider first the case where the stress-strain relation is linear, as shown in Fig. 65. A material having this property is called Hookean. The initial and transformed relations are
a=
F
a
E(r
AE
(e e0)
k(e a0)
(630)
F + e0 = fF + e0
We call k, f the stiffness and flexibility factors for the bar. Physically, k is the force required per unit elongation and f, which is the inverse of k, is the elongation due to a unit force. We consider next the case where the stress-strain relation is approximated by a series of straight line segments. The material is said to be piecewise linear. Figure 66 shows this idealization for two segments. A superscript (j) is used to identify the modulus and limiting stress for segment j. The forceelongation relation will still be linear, but now we have to determine what
SEC. 6-4.
127
segment the deformation corresponds to and also whether the strain is increasing (loading) or decreasing (unloading). For unloading, the curve is assumed to be parallel to the initial segment.* The relations for the various
possibilities are listed below.
1.
=
k">(e
F>1>
F=
2.
(63 1)
LoadingSecond Segment
F>l) < F
F
,4a12>
F>2>
k'2>(e
(632)
=
3.
+ (f>
UnloadingSecond Segment
k>1>(e
(633)
One can readily generalize these relations for the nth segment.f
*
We
are neglecting the Bauschinger etlect. See Ref. 2, Sec. 5.9. or Ref. 3, Art. 74.
128
CHAP. 6
Example 64
We consider a bilinear approximation, shown in Fig. E64.
Fig. E64
40
41.7
30
(in./in.)
Taking
L=lQft=l2Oin
we obtain
,4=lin.2
f"> =
f"> =
120 1/k"> =
24 x
=
k>2> =
L
AE>2>
= 83.3 kips/in.
12
in./kip
. = 41.7 kips/in. L
in./kip
F"> =
42>
= 3okips
+ ([1)
0.36 in.
Segment 1 Segment 2
F
F
(83.3)(e 120
(41.7)(e
Suppose a force of 35 kips is applied and the bar is unloaded. The equivalent initial
strain is (see Equation 633 and Fig. 66):
=
=
+ (f>2)
+ 0.06 in.
The procedure described above utilizes the segment stiffness, which can be interpreted as an average tangent stiffness for the segment. We have to modify the stiffness and equivalent initial elongation only when the limit of the seg-
SEC. 6-4.
129
is reached. An alternate procedure is based on using the initial linear stiffness for all the segments. In what follows, we outline the initial st(ffness approach.
ment
.4
'I I'
/1
-eo,eq.
segment 2 as
(634)
=
where e0,
eq
kU)(e
eo,eq)
is
(635)
The equivalent initial strain, eoeq, depends on e, the actual strain. Since e in turn depends on F, one has to iterate on eoeq regardless of whether the segment limit has been exceeded. This disadvantage is offset somewhat by the use
for all the segments. The notation introduced for the piecewise linear case is required in order to distinguish between the various segments and the two methods. Rather than continue with this detailed notation, which is too cumbersome, we will drop all the additional superscripts and write the force-deformation relations for bar n in the simple linear form
of
= =
(636)
130
CHAP. 6
where k, f, and e0 are defined by (631) through (635) for the physically
nonlinear case.
GENERAL BAR FORCEJOINT DISPLACEMENT RELATION
65.
The force-deformation and deformation-displacement relations for bar n are given by (622) and (636). Combining these two relations leads to an expres-
sion for the bar force in terms of the displacement matrices for the joints at the ends of the bar. The two forms are:
F,, =
and
e0,
F0,,,
k,,y,,u,,
(637)
F0,,, = k,,e0,,,
u,_) = e,,
e0,, + f,,F,,
(638)
We can express the force-displacement relations for the "m" bars as a single matrix equation by defining
(6 39)
k2
k1
k=
km
and noting (624). The generalized forms of (637) and (638) are:
F=
and
k(e
e0
e0)
F0
(640)
(641)
d1I =
66.
+ fF
Let F,, be the axial force vector for bar n (see Fig. 68). The force vector has the direction of the unit vector, i,,, which defines the orientation of the bar in the deformed state. Now, = fi,,i. Then,
F,, =
(642)
When F,, is positive, the sense of F,, is the same as the positive sense for the bar. Continuing, we define F,,,, as the forces exerted by bar n on the joints at the positive and negative ends of the bar. From Fig. 68,
= F,, = F,,fi,,i
F,,,,
+F,, =
(643
SEC. 66.
Joint n_
We consider next joint k. The external joint load vector is Pk, where For equilibrium, the resultant force vector must equal zero. Then, Pk =
Pk
- j+=k
The first summation involves the bars which are positive incident on joint k (positive end at joint k) and the second the bars which are negative incident. Using (6--43), the matrix equilibrium equation for joint k takes the form:
Pk =
Let
j+k
P2,
(644)
(if x
1)
(645)
(646)
Note that the rows of pertain to the joints and the columns to the bars. We partition into submatrices of order i x 1.
=
(if x
and
m)
1,
1,
2,.. ,j
.
k=
2.
,m
(647)
Since a bar is incident only on two joints, there will be only two elements in any column of From (644), we see that, for column n,
= =
(648)
e
when
orn_
The matrix can be readily developed using the connectivity table. It will have the same form as dT with y, replaced by n,,. When the geometry is linear,
and
132
CHAP. 6
Example 65
The
matrix for the truss of Example 61 has the following general form:
Bar Numbers
1
10
11
+ll
RT
T
2
nT
OF +p7
nT
plo
OT
ftT
I z
a
C
-'
A
DT
P2
oT 1J3
oT
nT
oT +p4 T
r
-I
J3
P6
oT +pU
I,
We could have also utilized the connectivity* matrix C to develop ft was pointed out in Example 63 that the elements of the kth column of C define the incidence of the bars on joint k. Using this property, we can write the
generalized form of (644) as
where
0
-.
(rn x Lm)
o
o
(649)
Finally, we have
=
67.
(650)
e = dl1 = e0 + IF =
and are the external joint-displacement and external joint-load matrices arranged in ascending order. Also, in our derivation, we
where the elements of
SEC. 67.
133
joint displacement restraints are imposed, there will be a reduction in the number of joint displacement unknowns and a corresponding increase in d, the number of force unknowns. This will require a rearrangement of and
when
Let r be the number of displacement restraints and 11d the number of displace-
ment unknowns. There will be n4 prescribed joint loads and r unknown joint loads (usually called reactions) corresponding to the na unknown joint displacenients and the r known joint displacements. We let U1, U2 be the column matrices of unknown and prescribed joint displacement components and P1, P2 the corresponding prescribed and unknown joint load matrices. The rearranged system joint displacement and joint load matrices are written as U, P:
(fld
x
><
1)
(r x 1)
651
We
1)
x 1)
+ V = 13
point out that the components contained in U (and P) may be referred to local reference frames at the various joints rather than to the basic frames. This is necessary when the restraint direction at a joint does not coincide with one of the directions of the basic frame. Finally, we let A and B be the transformation matrices associated with U and P. Then, (a) takes the form:
e=
P
BF
= AU
e0 + fF
(mxr)
A2]
(652)
A1U1 + A2tJ2 =
e0
+ fF
=
P2
B1F
B2F
Equation (653) represents equations relating the in unknown bar forces, the nd unknown displacements, and the r prescribed displacements. Equation
equations involving the in unknown bar forces and the prescribed joint loads. Lastly, Equation (655) represents r equations. for the r reactions in terms of the m bar forces. When the geometry is nonlinear, A and B involve the joint displacements. If the geometry is linear, A J3T, and
(654) represents
= AT
= 1, 2
(656)
134
CHAP. 6
We have introduced the displacement restraints into the formulation by with A, B. It remains to discuss how one determines A, B replacing d, In the following section, we treat the case of an arbitrary restraint from d,
direction. We also describe how one can represent the introduction of displacement restraints as a matrix transformation.
68.
When all the restraint directions are parallel to the direction of the global reference frame, we obtain U from by simply rearranging the rows of such that the elements in the first rows are the unknown displacements and the last r rows contain the prescribed displacements. To obtain A, we perform the same operations on the columns of d, Finally, since P corresponds to U, we obtain B by operating on the rows of or alternately, by operating on the columns and then transposing the resulting matrix. When the restraint at a joint does not coincide with one of the directions of the basic frame, it is necessary first to transform the joint displacement and external load components from the basic frame to a local frame associated with the restraint at the joint. Suppose there is a displacement restraint at joint k. Let (j = 1, 2, 3) be the orthogonal directions for the local reference frame associated with the displacement restraint at joint k. Also, let and be the corresponding displacement and external joint load components. Finally, let R0k be the rotation transformation matrix for the local frame at joint k with respect to the basic frame (frame o). The components are related by:
Uk
k_
ok
k.. Pk
where
ROk
Uk
ak
Pk
= [cos
(658)
We have omitted the frame superscript (o) for quantities referred to the basic frame (ut, to simplify the notation. We define CU', as the system joint-displacement and -force matrices referred to the local joint reference frames,
and
(6-59)
R1
=
Then,
R02
.
(660)
R0j
= =
(a)
SEC. 68.
135
(IP =
leads to
(661)
The transformation to is the same as for the case where the restraint directions are parallel to the directions of the basic frame, that is, it will involve only a rearrangement of the rows of Similarly, we obtain A by rearranging the columns of .cifi. The steps are
A2]
-+
-* B
LB2
Example 66
To obtain the submatrices in column k of
we postmultiply
k of ri by R"
resulting matrix or, alternately, we can premultiply the submatrices in row k of by R". As an illustration, see the matrix for Example 65 on page 136. The matrix can be and replacing il,, by y,. determined by transposing
One can visualize the introduction of displacement restraints as a matrix transformation. We represent the operations
U
as
and
(662)
U=DlI
and call D the displacement-restraint transformation matrix. When the restraint directions are parallel to the directions of the basic frame, D is a permutation matrix which rearranges the rows of We obtain D by applying the same row rearrangement to a unit matrix of order ij. Postmulti-
U =
(663).
See
C)
for Example 66
0
F4
F5
F6
F7
F1
F2
F8
F9
F3
F10
F11
Pt
R'pb
z z
0 C
pt
0 z C,)
m
R04P1
:13
> 2
ROSDTI
m
-1
(I)
Cl)
SEC. 69.
INITIAL INSTABILITY
137
U=
and it follows that
I) =
Since both H and
Using (662),
HPII'
(664)
A
and
dDT
then substituting for d, P.s, and D in terms of the geometrical, connectivity, local rotation matrices lead to
B
A=
( 6 65 )
Equation (665) is of interest since the various terms are isolated. However, one would not generate A, B with it.
69.
INITIAL INSTABILITY
The force equilibrium equations relating the prescribed external joint forces and the (internal) bar forces has been expressed as (see Equation 654):
P1=B1F
where P1 is
I) and F is (m x 1). When the geometry is nonlinear, B1 depends on the joint displacements as well as on the initial geometry and
<
restraint directions. In this section, we are concerned with the behavior under an infinitesimal loading. Since the nonlinear terms depend on the load intensity, we they will be negligible in comparison to the linear terms for this case, take B1 as constant. Then, (a) represents linear equations in in unknowns. If these equations are inconsistent for an arbitrary infinitesimal loading, we say the system is initially unstable. When the geometry is linear, B1 is independent of the loading and the initial stability criterion is also applicable for a finite loading. This is not true for a nonlinear system. We treat stability under a finite loading in Chapter 7. Consider a set of j linear algebraic equations in k unknowns.
ax=c
(b)
In general, (b) can be solved only if a and [a c] have the same rank,* It follows
that the equations are consistent for an arbitrary right-hand side only when
the rank of a is equal to], the total number of equations. Applying this condition
*
See
138
CHAP. 6
to (a), we see that the truss is initially unstable when the rank of B1 is less than na. For the truss to be initially stable under an arbitrary loading, B1 must be That is, the number of bars must be at least This requires m of rank
equal to the number of unknown displacement components. Since the rank this condition is necessary but not sufficient for initial may still be less than stability. In order to determine whether a truss is initially stable, one must actually find the rank of B1. The following examples illustrate various cases
of initial instability.
Example
67
x2
in =
4
5
na =
x1
F1
Pu
F F2
F3
F4
P12
P21
+1
B1
+1 +1 +1
P22
P31
Row 3 is ( 1) times row 1. The equations are consistent only if P21 = Pu Since m < we know the system is unstable for an arbitrary loading without actually finding r(B1).
SEC. 69.
INITIAL INSTAB1LITY
139
Example 68
We first develop the matrix for the truss shown in Fig. E68A and then specialize it for various restraint conditions.
Fig. E68A
'1
J
0
M
4
3
I
F6
cosU
F
F
F2
F3
F4
F5
Pit
Piz
Psi
P22
+1
sinO
+1
a---
1 +1
1
cosO
.
sinO
cosO
sinO
Psi
P32
P41
1
cosO sinO
row+row+row= row
row + row + row
= row
(sin 8)(row + row ) cos U (row ) = cos U (row )
(2) (3)
140
CHAP. 6
The first two relations correspond to the scalar force equilibrium conditions for the external
joint loads:
Pkl
P11
Pk2 = P12
k1
Mk is the moment of the external force vector acting at joint k with respect to point
0, the origin of the basic frame. We obtain relation (3) by taking Oat joint 4. Equation (b)
reduces to
d(p11 + P21) + b(p22
Using
ci
+ P32) =
=L
sin 8
b =
L cos
cos 0p32
sin
which is relation 3.
We see that rows 2 and 5 arc independent. Thc remaining set (rows 1, 3, 4, 6, 7, 8) contains only three independent rows. Now, we obtain B1 from by first taking a linear combination of the rows (when the restraints are not parallel to the basic frame) and then deleting the rows corresponding to the joint forces associated with the prescribed joint displacements. Since has three linear dependent rows, it follows that we must introduce at least three restraints. Initial instability will occur if
1.
2.
An insufficient number of restraints are introduced (n4 > 5). A sufficient number of restraints are introduced (/24 = 5) but the rows of B1 are not linearly independent. We say the restraints are not independent in this case. These cases are illustrated below.
Case
Fig. E68B
1
x2
m6
x1
SEC. 69.
INITIAL INSTABILITY
We obtain B1 by deleting rows 6 and 8 (corresponding to P32 and P42). The system is stable only when the applied joint loads satisfy the condition
Pu + P21 + P31 =
Case 2
P41
Fig. E68C
x2
rn = 6
5
xl
We delete rows 4, 6, and 8. The number of restraints is sufficient (fld = 5) but the restraints are not independent since r(B1) < 5. Actually, r(81) = 4. To make the system stable, at least onc horizontal restraint must be introduced.
In Example 68, we showed that there are three relations between the rows for a two-dimensional truss. These relations correspond to the force- and moment-equilibrium conditions for the complete truss. To establish the relations for the three-dimensional case, we start with the equilibrium equations, (jxl) 3
o
(2i3)x 1
is the moment of with respect to an arbitrary moment center, 0. For convenience, we take 0 at the origin of the basic reference frame. Parwhere
titioning
(666)
where
142
CHAP. 6
Sec. 52
=0
Equation (667) represents i relations between the rows of PA,
2)
+ q] = row [i(j
1)
q=l,2,...i
(669)
+ q]
We have shown that there are at least 3(i 1) relations between the tows of PA. Now, we obtain B by combining and rearranging the rows of PA. It follows that B will also have at least 3(i 1) relations between its rows. Finally, we obtain by deleting the rows corresponding to the restraints. For the system to be initially stable, we must introduce at least 3(i 1) restraints:
r
no.
of restraints
3(i
1)
(670)
Note that this requirement is independent of the number of bars. Also, it is a necessary but not sufficient condition for initial stability. in. The number of restraints must also satisfy the necessary condition This requires r = ((j m) (671) Both (670) and (671) must be satisfied. Either condition may control r, depending on the arrangement of the bars.
REFERENCES
C. H., and J. B. WILBUR: Elementary Structural Analysis, McGraw-Hill, New York, 1960. 2. CRANDALL, S. H,, and N. C. DAHL: An Introduction to the Mechanics of Solids, McGraw-Hill, New York, 1959. 3, TIMOSNENKO, S.: Strength of Materials, Part 2, Van Nostrand, New York. 1941. 4. TIMoSISENKO, S., and D. H. YOUNG: Theory of Structures, McGraw-Hill, New York,
1.
NORRIS,
1945.
5.
6. 7.
8.
MCMINN, S. I.: Matrices jbr Structural Analysis, Wiley, New York, 1962. MARTIN, H. C.: Introduction to Matrix Methods of Structural Analysis, McGrawHill, New York, 1966. LIVIISLEY, R. K.: Matrix Methods of Structural Analysis, Pergamon Press, London,
1964.
FENVES, S. J., and F. H. BRANIN: "Network-Topological Formulation of Structural Analysis," J. Struct. Div., ASCE, Vol. 89, No. ST4, pp. 483514, 1963.
Eq. 511.
* See
PROBLEMS
143
PROBLEMS
61.
(a)
(b)
62.
= {l,
1,0] (ft)
{5, 5. 2] (ft)
(a) (b)
Suppose
Uk
1/10,
1/30}
(inches) (inches)
Note that the units of x and u must be consistent. Find 1k and Determine and ji,,, using the exact expressions (Equations 615, 617), the expressions specialized for the case of small strain (Equations 619, and 620), and the expressions for the linear geometric case (Equation 621). Compare the results for the three cases. 63. Discuss when the linear geometric relations are valid and develop the appropriate nonlinear elongation-displacement relations for the trusses shown. Assume no support movements. 64. Consider the truss shown: (a) Establish the connectivity table.
(b)
(c)
144
CHAP. 6
Prob. 63
.3
(a)
x2
X1
3 2
Ib)
x2
Prob. 64
I'
(d) (e)
65.
Determine C.
Verify that d =
cxC.
66.
Determine d for the three-dimensional truss shown. Consider the d-c network shown. The Junctions are generally called
nodes, and the line connecting two nodes is called a branch. The encircled numbers refer to the branches and the arrowheads indicate the positive sense (of the current) for each branch.
PROBLEMS
.
145
Let and n. (j = 1, 2,. . , 5) denote the potential at node j. Also, let denote the nodes at the positive and negative ends of branch n. The potential
Prob. 65
x3
4
(0, 1,0)
xI
(1,0,0)
(1,1,0)
Prob. 66
3
0'
drop for branch n, indicated by
We define v and e as
is given by
= v,,,
v= e=
{v1,
v2,.. .
.
v5}
{e1, e2, .
e=
.cjv
Determine d, using the branch-node connectivity table. Discuss how the truss
problem differs from the electrical network problem with respect to the, form of ad. How many independent columns does ad have? In network theory, ad is called the augmented branch node incidence matrix.
146
CHAP. 6
67.
(a) (b)
(c)
Take L 20 ft, A = 2 in2, and the a-s curve shown. Develop the piecewise linear force-elongation relations. Suppose a force of + 60 kips is applied and then removed. I)etermine the force-elongation relation for the inelastic case. Suppose the bar experiences a temperature increase of 1000 F. Determine the initial elongation. Consider the material to be aluminum.
Prob. 67
6X
ksi
20 ksi
68.
and express eb0 in terms of quantities associated with segment (J 1). 69. Generalize Equation 635 for segmcntj.
610.
Suppose
as in the sketch, by
a=
E(s
be3)
Prob.
610
GA
Ee
da
Et
PROBLEMS
(a)
147
Determine expressions for ES and E', the secant and tangent moduli. Determine expressions for k5 and kt. Suppose the material behaves inelastically for decreasing 4 Consider the unloading curve to be parallel to the initial tangent. Determine the force-elongation relation for AB. 6li. Repeat Prob. 610, using the stress-strain relation
(b) (c)
=
where E, c, and n are constants.
612.
(u +
612
(a)
(b) Assemble for the linear geometric case. 613. Repeat Prob. 612 for the three-dimensional truss shown.
Prob.
613
LX2
'I.
is from, node
614.
(a)
Consider the electrical network of Prob. 66. Let be the current in branch n. The positive sense of
n_ to node n. Now, the total current flowing into a node must equal the total current flowing out of the node. This requirement leads to one equation for each node involving the branch currents incident on
148
CHAP. 6
the
node. Let
=
{i1, j2, .
.
(b)
where d is given in Prob. 66. How many independent equations does (a) represent?
(Hint: d has only four independent columns). When the resistance is linear, the current and potential drop for a branch are related by
=
e0
alternate form is
Note the similarity between (b) and the linear elastic member forceelongation relation. Show that the complete system of branch curpotential relations can be written as
e=
= e0
+ Ri
= R1(e
e0) = R1dv
R1e0
Equations (a) and (c) are the governing unpartitioned equations for a linear-resistance d-c network. The partitioned equations are developed in Prob. 623. It should be noted that the network problem is onedimensional, that is, it does not involve geometry. The d matrix depends only on the topology (connectivity) of the system. Actually, d corresponds to the C matrix used in Sec. 63 with i = 1. 615. Refer to Prob. 612, Suppose u11, u42, is52 are prescribed. Identify
B1 and B2. 616. Refer to Prob. 612.
(a) (b)
PROBLEMS
617.
(a) (b)
Develop the general form of Determine B1 and B2 corresponding to the following prescribed displacements:
U11,
U12,
U3j,
U-33,
U23,
U13
x1
x2
1/2 1/2
x3
0
1/2
1/2
648. Consider the two-dimensional truss shown. The bars are of equal length and 0 is the center of the circumscribed circle. The restraint direction degrees counterclockwise from the tangent at each joint. Investigate the is initial stability of this system. Repeat for the case of four bars.
Prob. 618
r (restraint direction)
t (tangent)
11
= 13
I.
619.
Suppose na =
0 are
in.
equations for P1
(mxl)
B1
F=0
Fk
rnXl
If (a) has a nontrivial solution, the rank of B1 is less than m and the system is initially unstable (see Prob. 145). Rather than operate on B1, to determine r(B1), we can proceed as follows: (1) We take the force in some bar, say bar k, equal to C:
150
CHAP. 6
(2)
(3) The last equilibrium equation leads to an expression for Fk in terms of C. If this reduces to an identity, r(B1) < since a nontrivial solution for
F exists. This procedure is called the zero load test.
(a)
Apply this procedure to Prob. 618. Take F1 = C and determine F2, F3, and then F1 using the equilibrium condition (summation of forces normal to r must equal zero) for joints 1, 2, 3.
When n4 = m and the geometry is linear, the truss is said to be statically determinate. In this case, we can determine F, using only the equations of static equilibrium, since the system, P1 = B1F, is square. Do initial
(b)
elongations and support settlements introduce forces in the bars of a statically determinate truss? 620. Modify the zero load test for the case where na < in. Note that the general solution of B1F 0 involves m r(B1) arbitrary constants. 621. Investigate the initial stability of the two-dimensional truss shown. Use the zero load test.
Prob.
621
622. Investigate the initial stability of the system shown. The restraint directions are indicated by the slashed lines.
Prob. 622
-j
4
-t
I
3
PROBLEMS
623. We generalize the results of Probs. 66. and 614 for a network having b branches and n nodes. Let
e
v
branch potential duff, matrix = {e1, e2, -. = branch current matrix = {i1, j2 = node potential matrix = {v1, v2,. . . ,
(nxb)
=
Ri
e = dv = e0 +
Now, dT
rows of
has
only n
are related by
independent equations, and one equation must be disregarded. Suppose we delete the last equation. This corresponds to deleting the last column of d (last row of dT). We partition d,
(bxn)
bx(n1)
bxl
d2]
and let d1 = A. The reduced system of node equations has the form
ATj
=0
Note that AT corresponds to B1 for the truss problem. Equation (e) represents (n 1) equations. Since v is of order n, one of the node potentials must be specified. That is, we can only determine the potential difference for the nodes with respect to an arbitrary node. We have deleted the last column of d which corresponds to node n. Therefore, we take as the reference potential.
(a)
Let
= {v1
v2
. . .
Show that
dv = AV
(b)
corresponds to introducing displacement restraints in the truss problem. Compare the necessary number of restraints required for the network and truss problems;
2. 3.
The elbngation of a bar was related to the translations of the joints at the end of the bar. Next, the bar force was expressed in terms of the elongation and then in terms of the joint translations. Finally, the equilibrium conditions for the joints were enforced, resulting in equations relating the external joint loads and internal bar
forces.
The system equations were obtained by generalizing the member forcedisplacement and joint force equilibrium equations and required defining only Later, in Chapter 10, we shall two additional transformation matrices follow essentially the same approach to establish the governing equations for an elastic solid. In this chapter, we develop two variational principles and illustrate their application to an ideal truss. The principle of virtual displacements is treated first. This principle is just an alternate statement of force equilibrium. Next, we discuss the principle of virtual forces and show that it is basically a geometrical compatibility relation. Both principles are then identified as the stationary requirements for certain functions. For this step, we utilize the material presented in Chapter 3, which treats relative extremas of a function. Finally, we discuss the question of stability of an elastic system and develop the stability criterion for an ideal truss. Why bother with variational principles when the derivation of the governing equations for an ideal truss is straightforward? Our objective in discussing them at this time is primarily to expose the ieader to this point of view. Also, we can illustrate these principles quite easily with the truss. Later, we shall
152
SEC. 72.
153
these principles, particularly the principle of virtual forces, to construct approximate formulations for a member.
use
w0 +
JFdv
= W(v)
where v0 is an arbitrary reference displacement. Since W is a function of v, the increment in W due to an increment Ar can be expressed in terms of the differentials of W when F is a continuous function of yr f = dW + 4d2W +
dW dW Ar = dv
F
Au
(72)
d2W = d(dW)
refer to dW as the first-order work. Similarly, we call d2W the secondIf dF/dv is discontinuous, as in inelastic behavior, we must use the value of dF/dv corresponding to the sense of Av. This is illustrated in
We
order work.
wrv0
154
CHAP 7
< 0. k2 for of v when there is a reversal in the Note that W is not a single-valued function
Fig. 72. We use dF/dv = +k1 for
F-v curve.
vs
We consider first a single mass particle subjected to a system of forces (see Fig. 73), Let R be the resultant force vector. By definition, the particle is in equilibrium when R = 0. We visualize the particle experiencing a displacement increment Au from the initial position. The first-order work is
dW=R'Afi
R=
(73)
If the initial position is an equilibrium position, dW 0 for arbitrary Au since 0. Therefore, an alternate statement of the equilibrium requirement is:
The first-order work is zero for an arbitrary displacement of a particle from an equilibrium position.
The incremental displacement
74
is called a virtual displacement; this statement is the definition of the principle of virtual displacements.
R+
One can readily generalize (74) for the case of S particles. Let
be the the
SEC. 72.
155
If particle q is in equilibrium,
for arbitrary It follows that the scalar force-equilibrium equations for the system are equivalent to the general requirement,
dW
for arbitrary
Equation (75) is the definition of the principle of virtual work for a system
of particles.
In general, some of the forces acting on the particles will be due to internal restraints. We define dW5 as the first-order work done by the external forces and dW1 as the work done by the internal restraint forces acting on the particles.
Substituting for dW, (75) becomes
dW5 + dW1 =
for arbitrary
S
q=l,2
be the work done by the internal restraint forces acting on Now. let the restraints. We use the subscript D for this term since it involves the
F1
/(Deformedl
I
.
F1
El
(Initial)
(Deformed)
F1
F1
II
Fig. 74. Work done on the mass particles and internal restraints.
magnitude, but opposite in sense, to the reaction of the particle on the restraint. Since the points of application coincide, it follows that
As an illustration, consider the simple system shown in Fig. 74. For this
case, we have
dW,, =
F1
Au1 + F1 Au2
F1 Au2
dW1 = F1 Au1
156
CHAP. 7
dWE =
dWD
for arbitrary
q=1,2,...,S
(76)
The first-order work done by the external forces is equal to the first-order work done by the internal forces acting on the restraints for any arbitrary virtual displacement of a system of particles from an
equilibrium position.
We emphasize again that (76) is just an alternate statement of the force equilibrium conditions for the system. Some authors refer to (76) as the
work equation. To apply the principle of virtual displacements to an ideal truss, we con-
sider the joints to be mass points and the bars to be internal restraints. We have defined and as the column matrices of external joint loads and
corresponding joint displacements, Then,
=
where &W contains the virtual joint displacements. The first-order work done
=
Generalizing (b), we have
F,, deH
dW0 = FraC
Finally, the work equation for an ideal truss has the form
Ml = FT de
(77)
The scalar force-equilibrium equations are obtained by substituting for de in terms of Mt. It is convenient to first establish the expression for the differential elongation of an individual bar and then assemble de. Operating on
u,
de, =
u,)T]
Au,,)
(Au
Au,)
(78)
1= J,,, F, de, = Wd(e,). We must use the rules for forming the differentials of a compound function since e, depends on the joint displacements. Using (317), we can write
dW4 =
d2
dWd
de
de,
= F, de,
(de,)2
= d(F, de,,) =
de,
+ F, d2e,
SEC. 72.
157
The assembled form follows from (625). We just have to replace y,, with il,,:
de
MI
(79)
?,TMI =
and requiring (a) to be satisfied for arbitrary
results in the joint force
force equilibrium equations. For the geometrically linear case, e where d is constant and de = dM1 follows directly from e. We have treated the geometrically nonlinear
case here to show that the principle of virtual displacements leads to forceequilibrium equations which are consistent with the geometrical assumptions associated with the deformation-displacement relations.
Example
71
We consider a rigid member subjected to a prescribed force, P. and reactions R,, R2, as in the diagram. There is no internal work since the body is rigid. Introducing the virtual
Fig.
E71
dW =
Now,
= R, Au1 + R2 Au2
is not independent:
= Au,
Then,
d\
+ Au2
7d
dW =
Au2 {R2
=0
R2 = P
which are the force and moment equilibrium equations, in that order.
158
CHAP. 7
Example 72
We consider the outside bars to be rigid (see sketch). To obtain the force equilibrium equation relating P and the internal bar forccs F1, F2, we introduce a virtual displacement, Au1, of the point of application of F:
de2 =
Au2 cos El
Au1 cos U
where U defines the initial position. Then, equating dW5 and dW0,
dWE =
(1WD
P=
(F1
Fjcos
The force in bar 3 does not appear explicitly in the equilibrium equation, (c). It is possible
Fig. E72
Now,
Au1 Au2
Multiplying the constraint relation by 2, adding the result to (d), and collecting terms, we obtain Au1(P F1 cos 0 2) + Au2(F2 cos 0 + 2) = 0 (f) Finally, we require (f) to be satisfied for arbitrary Au1 and Au2. The equilibrium equations
See See, 33.
159
P=F1cosO+A F2cos6+2=O
and we recognize 2 as the force in bar 3.
73.
The principle of virtual forces is basically an alternate statement of geometrical compatibility. We develop it here by operating on the elongation joint displacement relations. Later, in Chapter 10, we generalize the principle. for a three-dimensional solid and describe an alternate derivation. We restrict this discussion to geometric linearity. The governing equations
= which satisfy
=
A force system which satisfies the equations of static equilibrium is said to be statically permissible. Equation (b) relates the actual elongations and joint dissum over the bars, and placements. If we multiply the equation for Ck by note (c), we obtain the result AFTe = =
which is the definition of the principle of virtual forces:
=0
(7-10)
for any statically permissible system of bar forces and joint loads.
where U2 contains the prescribed support movements. Using (a), (710) takes the form: AFTe U2 = APr U1
160
CHAP. 7
If the elongations are known, we can determine the unknown displacements by specializing AP1. To determine a particular displacement component, say we generate a force system consisting of a unit value of PkJ and a set of bar forces and reactions which equilibrate Pkj = 1.
= (711) = The internal bar forces and reactions are obtain from an equilibrium analysis of a statically determinate structure. Since only one element of is finite,
APf U1
(l)ukj
(712)
The principle of virtual forces is also used to establish geometric compatibility relations required in the force method which is discussed in Chapters 9 and 17. We outline the approach here for completeness. One works with
self-equilibrating virtual-force systems, i.e., statically permissible force systems which involve only bar forces and reactions. By definition, a self-equilibrating force system F*, f)* satisfies
B1F* =
For this case, (b) reduces to
.
P7 = P7 = B2F*
(713)
(714)
Example
73
The truss shown (Fig. E73A) has support movements and is subjccted to a loading which results in elongations (e1, e7) in the diagonal bars. We are coniidering the outside bars to be rigid.
Fig. E73A
u,P
SEC. 73.
To determine the translation, u, we select a statically determinate force system consisting of a unit force in the direction of u and a set of bar forces and reactions required to equilibrate the force. One possible choice is shown in Fig. E7.-3B. Evaluating (712) leads to
u
cos 6
tan
This truss is statically indeterminate to the first degree. A convenient choice of force redundant is one of the diagonal bar forces, say F2. The equation which determines F2 is
0
Fig. E73B
derived from the gcometric compatibility relation, which, in turn, is obtained by taking a self-equilibrating force system consisting ofF2 = + I and a set of bar forces and reactions required for equilibrium. The forces are shown in Fig. E7--3C.
Fig. E73C
ens 6
0
Evaluating (714), we obtain
e1 + e2 = 0
To show that (a) represents a geometrical compatibility requirement, we note that the elongation-displacement relations for the diagonal bars are
ucos6
e2
= ucosO
Specifying e1 determines u and also e2. We could have arrived at Equation (a) starting
from Equation (b) rather than (714). Flowever, (714) is more convenient since it does not involve any algebraic manipulation. We discuss this topic in depth later in Chapter 9.
162
CHAP. 7
74.
In this section, we specialize the principle of virtual displacements for elastic behavior and establish from it a variational principle for the joint displacements. We start with the general form developed in Sec. 72,
FT
de =
gpT
de =
A'W
and (a) leads to the complete set of force-equilibrium equations in unpartitioned form,
=
can obtain the equation for P1 by rearranging (c) or by starting with the partitioned form of
We
iW1 =
(715)
where now
In what follows, we will work with (715).
=
we can write (715) as
VT
= for arbitrary
(717)
(718)
We call the total strain energy function and the total potential energy. One should note that VT exists only when F is a continuous single-valued function of e. This requirement is satisfied when the material is elastic. Equation (718) states that the joint force-equilibrium equations (P1 = B1F) expressed in terms of the unknown displacements are the Euler equations for the
SEC. 74.
STRAIN ENERGY
163
total potential energy. It follows that the actual displacements, i.e., the displacements which satisfy the equilibrium equations, correspond to a stationary
value of 11,'
and
=
where is the strain energy for bar). Since we are considering pound function of e1, Equation (b) is equivalent to
to be a com(719)
= That is, the strain energy function for a bar has the property that its derivative
with respect to the elongation is the bar force expressed in terms of the elongation. Finally, we can express as
(720) = where e0 is the initial elongation, i.e., the elongation not associated with the force. Actually, the lower limit can be taken arbitrarily. This choice corresponds to taking as the area between the F-c curve and thee axis, as shown in Fig.75.
Then
=
VTas
VT
e0,
)2
(721)
The total strain energy is obtained by summing over the bars. We can express
=
j=1
4(e
e0)Tk(e
eo)
(722)
164
CHAP. 7
e=
A1U1 + A2U2
(723)
When the geometry is linear, A1, A2 are constant and is a quadratic function. If the geometry is nonlinear, is a fourth degree function of the displacements. Up to this point, we have shown that the displacements defining an equilibrium position correspond to a stationary value of the potential energy function. To determine the character (relative maximum, relative minimum, indifferent,
neutral) of the stationary point, we must examine the behavior of the second differential, in the neighborhood of the stationary point. and noting that AP1 = 0 leads to Operating on
d2
(724)
The next step involves expressing d2VT as a quadratic form in AU1. We restrict this discussion to linear behavior (both physical and geometrical). The general
nonlinear case is discussed in Sec. 17.6 When the geometry is linear, we can operate directly on (723) to generate the differentials of e,
de
A1AU1
0
d2e =
dF=kde
where k is a diagonal matrix containing the stiffness factors (AE/L) for the bars. Then, d2VT reduces to
d2VT = dFT dc = deTk de
AUT(ATkA1)AU1
7 25
If de
0 for all nontrivial AU1, d2VT is positive definite and the stationary
0 for AU1 0
A1 AU1 =
(m equations in
have a nontrivial solution. But a nontrivial solution of (a) is possible only when for the geometrically linear case and r(B1) = r(A1) < ne,. However, A1 = when the system is initially stable. Therefore, it follows that the displacements
Example
74
We establish the total potential energy function for the truss considered in Example 72.
SEC. 75.
energy is
COMPLEMENTARY ENERGY
165
VT
e1 =u1cosO
results in
=
and finally
=
The first differential of
is
P1}Au1
P1 =
P1 =
F1 =
=
(F1
F2)cos 0
ku1
cos 0
F2 =
k2e2
k2u1
cos 0
is
=
and
P1
= k1 +
0. Thc truss is initially unstable
75.
The principle of virtual forces can be transformed to a variational principle for the force redundants. We describe in this section how one effects the trans-
eTAF
where AF,
represent a self-equilibrating force- system, i.e., they satisfy the the following constraint relations:
B1 iW = 0
B2 AF
166
CHAP. 7
Our objective is to establish a function of F, whose Euler equations are (a) and
(b). We cannot work only with (a) since F is not arbitrary but is constrained by the force-equilibrium equations,
P1 =
We interpret
B1F
(fld
equations in m variables)
=
and call
eT
dV7
(726)
(727)
That is, the complementary energy function for a bar has the property that its derivative with respect to the bar force is the elongation expressed in terms of the force. We express as
dF1
(728)
This definition corresponds to taking Vj' as the area bounded by the F-e curve
and the F axis as shown in Fig. 75. Also, the strain and complementary
energy functions are related by
+
When the material is linear elastic,
=
+
(729)
=
11
e0,
1.1r2 rj T 2Jj' 3
+ 4FTfF
731)
=
We call
the total complementary energy function. With these definitions, Equations (a), (b), and (c) can be interpreted as
0
We can combine (e) and (f) into a single equation by introducing Lagrange multipliers. Following the procedure described in Sec. 33, we add to (731) the joint force equilibrium equations and write the result as:
+ (P1
where
.
(732)
SEC. 75.
COMPLEMENTARY ENERGY
167
tions for
treating F and
dfl, =
for AF,
arbitrary
(733)
e(F) B1F = P1
+ B102
We recognize the first equation in (733) as the member force-displacement relation, and it follows that = U1.
bar forces. One can also work with a combination of bar forces and reactions as force unknowns. We let
(rn
q= m X = {X1, X2
(734)
F=
P2
+ P2,0 +
F0
is
(735)
self-equilibrating, i.e.,
for arbitrary X
to
(736)
Then,
AF
UI AP2
UIP2,
x
=
and the Euler equations are
eTFx
(eTFx
UIP2
(737)
Note that (737) is just a reduced form of (733). Also, we could have obtained this result by substituting directly in (a).
Up to this point, we have shown that the force redundants which satisfy the geometric compatibility equations correspond to a stationary value of the total complementary energy. To investigate the character of the stationary point, we evaluate the second differential. Operating on (g),
deTFx AX
d2 is positive definite with regard to AX, the stationary point is a relative minimum. This requirement is satisfied for the linear elastic case. To show this, we note that
de
= fAF = WXAX
=
168
CHAP. 7
Since f contains only positive elements, AX provided that there does not exist a
AX =
=0
For (j) to have a nontrivial solution, there must be at least one relation between the columns of But this would correspond to taking force redundants which
are not independent, and the solution scheme would degenerate. Therefore, we can state that the actual force redundants correspond to an absolute minifor the linear elastic case. mum value of
Example
75
We consider the truss treated in Example 73. It is statically indeterminate to the first degree with respect to the bars (statically detcrminate with respect to the reactions) and we take X = F2
The force influence matrices defined by (735) follow from the force results listed on the sketches:
F0 =
0; 0; 0; 0; tan 0; 0}
= {+1; +1; cosU; sin 0; sin 0; cos0} P20 = P{l; tanG; +(an0} =0
Assuming a bar is rigid is equivalent to setting f = mentary energy is due only to the diagonal bars:
= yr + = e02F2 +
We convert
+ f2Fi')
to a function of X by substituting
F1 =
Finally,
has the form
cos 0
F2 = +X
=
tan 0)P + + 4(f1 + f2)X2
e0,1
tan 0
(eoi + e0,2 + f1
{[eoi +
(fl
d211, = (f1 +
(g)
SEC. 76.
STABILITY CRITERIA
(f1
+ f2)X = 0
is the
Comparing (h) with (a) of Example 73, we see that the Euler equation for geometric compatibility equation expressed in terms of the force redundant.
76.
STABILITY CRITERIA
Section 69 dealt with initial stability, i.e., stability of a system under infinitesimal load. We showed there that initial stability is related to rigid body motion. A system is said to be initially unstable when the displacement restraints are insufficient to prevent rigid body motion. In this section, we develop criteria for stability of a system under finite loading. If a linear system is initially stable, it is also stable under a finite loading. However, a nonlinear (either physical or geometrical) system can become unstable under a finite load. We consider first a single mass particle subjected to a system of forces which are in equilibrium. Let Il be the displacement vector defining the equilibrium position. We introduce a differential displacement All, and let AW he the work done by the forces during the displacement All. if A W > 0, the particle energy is increased and motion would ensue. It follows that the equilibrium position (ll) is stable only when AW < 0 for arbitrary All. We consider next a system of particles interconnected by internal restraints. Let AWE be the incremental work done by the external forces and AW1 the incremental work done by the internal restraint forces acting on the particles. The total work, AW, is given by
AWE + AW1
The system is stable when A W < (I for all arbitrary permissible displacement increments, that is, for arbitrary increments of the variable displacements. Now, we let AW11 be the work done by the internal restraint forces acting on the restraints. Since AW1, we can express the stability requirement as (7-38)
One can interpret AW0 as the work required to deform the system to the alternate
position and as the actual work done on the system. When the behavior is continuous, we can express and AWE as Taylor series expansions in terms of the displacement increments (see (72)):
dW0
+ +
0
+ +
170
CHAP. 7
If we retain only the first two terms in (b), the general stability condition reduces to
d2 WD d2 W5
>0
Equation (739) is called the "classical stability criterion." Retaining only the first two differentials corresponds to considering only infinitesimal displacement increments. If (739) is satisfied, the equilibrium position is stable with respect to an infinitesimal disturbance. In order to determine whether it is stable with respect to a finite disturbance, one must use (738). If
d2WD = d2WE
(740)
for a particular set of displacement increments, the equilibrium position is said position infinitesimally to be neutral, and there exists an alternate displaced from the first position. One can interpret (740) as the necessary condition for a bifurcation of equilibrium positions. To show this, suppose U and U represent the displacement components for the two possible equilibrium positions of a system where
= R + dR +
d2W
Now, the second-order work for the initial equilibrium position is given by
d2W5
d2WD
txUT dR
If d2 W =
it follows that
dR = R0AU =
The condition
to he infinitesimal,
R=R+dR
and (740) implies R =
0.
To apply the classical stability criterion to an ideal truss, we note that the
first-order work terms have the form
dW5 =
dWD =
P1
AU1
>jFjd2ej +
(7-41)
SEC. 76.
STABILITY CRITERIA
for arbitrary nontrivial AU1 d2 WD = 0 for a particular nontrivial AU1 (742) d2 WD < 0 for a particular nontrivial AU1 where d2WD is a quadratic form in AU1. We postpone discussing how one transforms (741) to a quadratic form in AU1 until the next chapter. When the material is elastic, we can identify (739) as the requirement that Fir, be a relative minimum. By definition,
d2WD > 0
drIp = dVT
dWE
stable) if it corresponds to a relative minimum (neutral, indifferent) stationary point of the total potential energy.
Example 76 The system shown in Fig. E76A consists of a rigid bar restrained by a linear elastic
spring which can translate freely in the .x2 direction. Points A and A' denote the initial and deformed positions. We will first employ the principle of virtual displacements to establish the equilibrium relations and then investigate the stability of the system.
Fig. E76A
x1
j2
P2 du2
(a)
172
CHAP. 7
where F, e are the spring force and extension. Since the bar is rigid, the system has only one
degree of freedom, i.e., only one displacement measure is required to define the configuration. It is convenient to take 0 as the displacement measure. The deformation-displacement relations follow from the sketch:
e=
u1
= L(sin
0 sin
0)
0 sin
and
de =
(cos 0)L
du, =
(sin 0)L
Using (a) and (d), the principle of virtual displacements takes the form
dW0 dWa
= {F cos 0
P2
sin 0) (L AO) =
for arbitrary AG
F cos 0 =
P2
sin 0
which is just the moment equilibrium condition with respect to point 0. We transform (1) to an equation for ()by substituting for F using (c). The result is
sin 0
tan
0=
sin 00
dW5
for arbitrary AU
The potential energy function for this system has the form
=
and (g) can be interpreted as
P2u2
= 4kL2(sin 0
sin
0)
=
0
for
0=
cos 0 = P2/kl
To investigate the stability of an equilibrium position, we have to evaluate the secondorder work at the position. After some algebraic manipulation, we obtain =
d2W0
= k(L
AU)2
[cossU P2/kL]
cos 0
Let 0* represent a solution of(g). Applying (m) to 0* results in the following classification:
REFERENCES
stable
COS 0* >
cos3
3 cos 0'
P2
neutral
unstable
P2
neutral unstable
dP2
dO
dP2
A transition from stable to unstable equilibrium occurs at point A, the peak of the deflection curve. The solution for 0 is different in that its stable segment is the linear kL) corresponds to a branch point, solution and the neutral equilibrium point (P2 Both the linear and nonlinear branches are unstable.
Fig. E76B
REFERENCES
1
2. 3. 4. 5.
6. 7.
WANG, C. T.: Applied Elasticity, McGraw-Hill, New York, 1953. LANGHAAR, H. L,: Energy Methods in Applied Mechanics, Wiley, New York, 1962. REISSNeR, E.: "On a Variational Theorem in Elasticity," J. Math. Phys., Vol. 29, pages 90-95, 1950. ARGYRIS, i. H., and S. KIsLseY: Energy Theorems and Structural Analysis, Butterworths, London, 1960. CI.IARLTON, T. M.: Energy Principles in Applied Statics, Blackie, London, 1959. HOFF, N. J.: The Anal vsis of Structures, Wiley, & New York, 1956. K.: Variational Methods in Elasticity and Plasticity, Pergamon Press, 1968.
174
CHAP. 7
PROBLEMS
71.
Consider the two-dimensional symmetrical truss shown. Assume = 03 = 0. (a) Determine the first two differentials of e1 and ez by operating on the
expanded expression (equation 619) for e.
(b)
(c)
When a b, we can neglect the nonlinear term involving u12 in the Specialize (a) for this case. expressions for e and When a b, we can neglect the nonlinear term involving u11 in the Specialize (a) for this case. expressions for e and
Prob. 71
x2
T
3
21
72. Refer to the figure of Prob. 71. Assume = u3 = 0 and a> b. Using the principle of virtual displacements, determine the scalar force-equilibrium equations for joint 1. 73. Suppose a force F is expressed in terms of e,
F=
a
C1e
+ 4C2e3
u + 1u2
(a)
=
C1(e
F de
(b)
F=
74.
e decreasing from e*. Determine d2 W at a = e*. Refer to Prob. 623. The n 1 independent node equations relating the branch currents are represented by ATI
U
1
Now, the branch potential differences, e, are related to the n node potentials, V, by
independent
e=
AV
PROBLEMS
175
for arbitrary
is equivalent to (a). Compare this principle with the principle of virtual displacements for an ideal truss. 75. Consider the two-dimensional truss shown. Assume u2 = = 0.
(a)
(b)
Using (714), obtain a relation between the elongations and 32. Take the virtual-force system as LxF2 and the necessary bar forces and reactions required to equilibrate AF2. Using (712), express u11, u12 in terms of e1, e3. Note that bar 2 is
x2
76. Refer to Prob. 623. One can develop a variational principle similar to the principle of virtual forces by operating on the branch potential differencenode potential relations. Show that
-
AiTe=0
for any permissible set of current increments. Note that the currents must satisfy the node equations
ATi
Deduce Kirchhoff's law (the sum of the voltage drops around a closed loop must equal zero) by suitably specializing Lsi in (a). Illustrate for the circuit
shown in Prob. 66, using branches 1, 2, 4, and 6. 77. By definition, the first differential of the strain-energy function due to an increment in U1 has the form
n1
dV.
F,, de,,
We work with expressed as a compound function of e = e(U) since it is more convenient than expressing V directly in terms of U1. One can also
176
CHAP 7
write
(a)
as
Using (b), show that the system of if joint force-equilibrium equations expressed in terms of the joint displacements can be written as:
U(k
k=
P(k=
Note that (d) is just the expansion of (c). Rework P rob. 72, using (d).
78. Determine V(e), dv, and d2V for the case where the stress-strain
relation has the form (see Prob. 610)
a=
be3)
79. Determine V*(F), dV*, and d2V* for the case where the stress-strain relation has the form
(a
ca3)
=
0Pkj
is defined by (731). This result specialized for U2 = 0 is = called Castigliano's principle, part H. Apply it to Prob. 75, part b. Assume linear elastic material and f1 = = = f. 711. The current and potential drop for a linear resistance are related by
where
ef
e0,j
as a function of e1.
=
(a)
=
Determine
b
corresponding to (b).
(b)
Let W
the branch potential drops to be functions of the node potentials, deduce that the actual node potentials V correspond to a stationary value of W. Use the results of Prob. 74. The Euler equations for
PROBLEMS
177
=
Determine
b
(d)
corresponding to (a).
(d)
Let W* =
H=
(e)
iTe
1T(AV)
= H(i,V)
(e)
are the governing equations for a d-c network. Show that the actual currents correspond to a stationary value of One can either introduce the constraint condition, An = 0, in (e) or use the result of Prob. 76. 712. Investigate the stability of the system shown below. Take k, = aL2k5
P
Prob.
712
Rigid rod
ICr
and
GENERAL
The basic equations defining the behavior of an ideal truss consist of forceequilibrium equations and force-displacement relations. One can reduce the system to a set of equations involving only the unknown joint displacements by substituting the force-displacement relations into the force-equilibrium equations. This particular method of solution is called the displacement or method. Alternatively, one can, by eliminating the displacements, reduce the governing equations to a set of equations involving certain bar forces. The latter procedure is referred to as the ,fin'ce or flexibility method. We emphasize that these two methods are just alternate procedures for solving the same basic equations. The displacement method is easier to automate than the force method and has a wider range of application, However, it is a computer-based method, i.e., it is not suited for hand computation. In contrast, the force method is more suited to hand computation than to machine computation. In what follows, we first develop the equations for the displacement method by operating on the governing equations expressed in partitioned form. We then describe a procedure for assembling the necessary system matrices using
only the connectivity table. This procedure follows naturally if one first operates on the unpartitioned equations and then introduces the displacement restraints.
The remaining portion of the chapter is devoted to the treatment of nonlinear behavior. We outline an incremental analysis procedure, apply the classical stability criterion, and finally, discuss linearized stability analysis.
82.
The
SEC. 82.
179
= B1F P2 = B2F
P1
eqs.)
The unknowns are the in bar forces (F), the r reactions (P2), and the na joint
displacements (U1). One can consider F, to represent the initial bar forces, that is, the bar forces due to the initial elongations and support movements 0. The term kA1U1 represents the bar forces due to U1. When the with U1 material is linear elastic, k and e0 are constant. Also, = BT when the geometry
is linear.
We obtain a set of equations relating the flj displacement unknowns, U1, by substituting for F in (a). The resulting matrix equation has the form
(B1kA1)U1 =
B1F1
We solve (8i) for U1, determine F from (e), and P2 from (b). The coefficient matrix for U1 is called the system stiffness matrix and written as
K11 = B1kA1
(82)
as representing the initial joint forces due to the initial One can interpret elongations and support movements with U1 = 0. Then B1F1 represents the net unbalanced joint forces. When the geometry is linear, K1 reduces to
1
If the material is linear, k is constant and positive definite for real materials.
Then, the stiffness matrix for the linear case is posiLive definite when the system
is initially stable, that is, when r(B1) Conversely, if it is not positive definite, the system is initially unstable. If the material is nonlinear, k and e0 depend on e. We have employed a piecewise linear representation for the force-elongation curve which results in linear relations. However, one has to iterate when the limiting elongation
for a segment is exceeded. The geometrically nonlinear case is more difficult since both A and B depend
on U1. One can iterate on (81), but this requires solving a nonsymmetrical system of equations. It is more efficient to transform (81) to a symmetrical system by transferring some nonlinear terms to the right-hand side. Nonlinear analysis procedures are treated in Sec. 84. Even when the behavior is completely linear, the procedure outlined above for generating the system matrices is not efficient for a large structure, since
f See Prob. 214.
180
CHAP. 8
it requires the multiplication of large sparse matrices. For example, one obtains the system stiffness matrix by evaluating the triple matrix product,
=AfkA1
One can take account of symmetry and the fact that k is diagonal, but A1 is generally quite sparse. Therefore, what is needed is a method of generating K which does not involve multiplication of large sparse matrices, A method which has proven to be extremely efficient is described in the next section.
83.
= F0,
=
where n4, n. denote the joints at the positive and negative ends of barn. One can consider F0, as the bar force due to the initial elongation with the ends
p,I_
= =
(84)
Substituting for
(84) expands to
=
pn_ =
One can interpret (b) as end actionjoint displacement relations since the elements of are the components of the bar force with respect to the
Note that is of order i x i where I = 2 or 3 for a two or three-dimensional truss, respectively. When the geometry is linear, and is sym= y,, metrical. With this notation, (b) takes a more compact form,
= =
(86)
We refer to as the bar stiffness matrix. Equation (86) defines the joint forces required for bar n. The total joint forces required are obtained by summing over the bars.
SEC. 83.
We have defined
p2, .
{u1, u2
X 1) (ii x 1)
(U
as the general external joint force and joint displacement matrices. Now, we write the complete system of if joint force-equilibrium equations, expressed in terms of the displacements, as
(87)
We refer to if, which is of order if x as the unrestrained system stiffness matrix. The elements of are the required joint forces due to the initial elongations and represents the required joint forces due to the joint
displacements.
We assemble if and in partitioned form, working with successive members. The contributions for member n follow directly from (86).
(Partitioned Form is j x 1)
in row
in row n.
88
if (Partitioned Form is j x j)
+k,
k,,
ku
Example 81
Fig. E8l
0
4
3
182
CHAP. 8
Bar
2
2
1
4 4
3
+joint
joint
2
3
U1
U2
U3
U4
k1+k2
P2
k2
k1
k2
k2 + k3 + k5
--k3
k5
k3
k3 + k4
k4
k4
k1
k1 + k4 + k0
Pai
,10
P0.2
Po, 3
L'
L'
0.1PI
0,212 r
0. 0,
UT
pT
L'
T r0,SPS
L'
3P3
1I'I 1
UT
t'
4P4
UT
Po, 4
uT 0, 41'4
0. 5P5
Example 82
The external force matrix, involves and the displacement matrices for those joints connected to joint j by bars. Now, corresponds to row j and ii,, to column j of ir. By suitably numbering the joints, one can restrict the finite elements of X' to a zone about the diagonal. This is quite desirable from a computational point of view.
Fig. E82
Sect. 1
0
(-I
-..- --
fs
(71
ft 2:
/
L6
',.
%._ --
-.
Consider the structure shown. We group the vertical joints into sections. The equilibrium equations for section k involve only the joints in section k and the adjacent sections. For example, the equations for section 3 (which correspond to P6) will involve only the displacement matrices for sections 2, 3, 4. This suggests that we number the joints by section. The unpartitioned stiffness matrix corresponding to the above numbering scheme
SEC. 83.
is listed below. Note that has the form of a quasi-tridiagonal band matrix when it is partitioned according to sections rather than individual joints. The submatrices for this truss are of order 4 x 4,
U'
U2
U3
U4
U6
U7
U8
k, +k2 k,
P2
Ik2
k4
k,
k,+k3 k3
+k4
k2+k3
k2
k3
k5
I
k6
k7
+k6
k4
k,
k4+k51
k8
k6
k6+k9 k9
k10
+k,0
p6 k7
k8
I
k9
k8+k7
+k9
k1,
k,2
+k,2
JJ7
k,0
k,,
k10+k,, k,,
3
p8
k,,
k,3
k,2+k,,
The introduction of displacement restraints involves first transforming the partitioned elements and to local frames associated with the restraints, permuting the actual rows, and finally partitioning the actual rows. The steps are indicated below.
-+ U
-+
joint frames as
+
and
T
(810)
= =
,n= 1,2,...,j
184
CHAP. 8
The step,
P,
We
obtain the corresponding stiffness matrix, K, by performing the same operations on both the rows and columns of The rearranged system of equations is written as P = KU + P0 (8-12)
Finally, we express (812) in partitioned form:
= P2 =
K11U1
13
Example
83
It is of interest to express the partitioned elements of K in terms of the geometrical, We start with the general Unconnectivity, and displacement transformation
partitioned equations(628), (640), and (644), (650):
F0 + kda/1 = F0 + kyC'W
=
=
Then, substituting for F in (a) and equating the result to (87) leads to
The matrix, DTkY, is a quasi-diagonal matrix of order im. The diagonal submatrices arc of order i, and the submatrix at location n has the form, We have defined this product as k,,. Then, if we let
=
k5
[ki
I. r,T
2p272
we can express
as
= CTk5c
Carrying out (89) for n = 1, 2 m is the same as evaluating the triple matrix product. Obviously, (89) is more efficient than (f).
P=
11
= =
and
D1dP
(g)
= DTU = DfU1 +
(h)
SEC. 83.
185
Substituting (g) and (h) in (87) and equating the result to (813), we obtain
K,, =
P0.
DsCTDke0
t12
then partition. This operation is quite time-consuming. Also, it leads to rectangular submatrices. In what follows, we describe a procedure for introducing displacement restraints which avoids these difficulties. We start with the complete system of equations referred to the basic frame,
and
(814)
We assemble
and
the
external force matrices for those joints which are unrestrained. It remains to modify the rows and columns corresponding to joints which are either fully or partially restrained.
Case A: Fit!! Restraint
Suppose uq = Uq. Then
On X. Set off diagonal matrix elements in row q and column q equal 1. to 0 and the diagonal matrix element equal to I,.
I=
(815)
= Ii
2.
On
Add terms in
t
due to
C
X(qUq
(816)
j
ease B: Partial RestraintLocal Frame
is
prescribed.
= prescribed = = unknown
186
CHAP. 8
4=
Step I Assemblage of Basic Matrices
G=O,
and we set
G,r
+1
2. us'. We start with an ith-order column vector having zero elements and we set the element in the rth row equal to Note that this matrix involves only the prescribed displacements (local frame) in their natural locations. We start with an ith-order column vector having zero elements and 3. we insert the values of the prescribed joint forces (local frame) in their natural locations. Note that the elements corresponding to the reactions are zero.
E=O,
Suppose joint
G=11
The rotation matrix, R5, defining the direction of the local frame at 5 with respect to the basic frame. (b) The direction (or directions) of the displacement restraint and the value (or values) of the prescribed displacement.
(a)
direction r,
(c)
j=1,...,i
As an illustration, suppose r =
In (c), we read in
5
2.
Psi
P53
2)
[0
0
1
E5=
0 0
0 0
0
1
Gs=IO
[0 0 0
SEC. 83.
187
In
by
e
=
" N, q
2.
'-'q 1%
" N, q
1' 2'
=
3.
irtq =
4.
1,
2, .
Add Gq to irqq
= it"qq + Gq
5.
Add
and
to
P'N,q + U
On Jr
= X'qq =
Oiz
2PN,
.YV'eq(E9R0")T
(Eq R").Y(qq(Eq
1'
R0q,
I?'N,q
1, 2,. . . ,j
(818)
+0
When ir is symmetrical (this will be the case when the system is geometrically linear), we can work only with the submatrices on and above the diagonal. The contracted operations for, the symmetrical case are threefold:
it"eq(EqR")T
(819)
= 1,2,...,q. 1
188
q
CHAP. 8
it'qq =
+ Gq
(820)
(821)
The operations outlined above are carried out for each restrained joint. Note that the modifications for joint q involve only row q and column q. We denote the mOdified system of equations by
(822)
will be Equation (822) represents if equations. The coefficient matrix nonsingular when K1 is nonsingular. To show this, we start with the first equation in (813) and an additional set of r dummy equations:
1
[K11
-fJ
Olfuil
fP0,1
Equation (a) represents 1/equations. This system is transformed to (822) when to d?tJ, we permute U, They are related by (sec (663))
U
[1?, J
= rVp
= HT[K11
IK11I
(823)
It is more convenient to work with (822) rather than (a) since the solution of (822) yields the joint displacement matrices listed in their natural order, that is, according to increasing joint number. Once ciii' is known, we convert the joint displacement matrices to the basic frame, using
uq =
F,, = F0,,, +
Next, we calculate F,,
and assemble
SEC. 83.
189
in row n+ in row n_
Once is known, we convert the force matrix for each partially restrained joint to the local joint reference frame, using
=
required to equilibrate the bar forces. This operation The final result is provides a static check on the solution in addition to furnishing the reactions. When the problem is geometrically nonlinear, y,, and depend on the joint displacements. In this case, it is generally more efficient to apply an incremental formulation rather than iterate on (822).
Example
84
/
Ii
1.
/50
2
3
1
4
3
6
3
7 5 3
9 5
10
11
5
+joint(n+)
joint(n...)
4
2
6 4
2.
Assemblage of =
and
Applying
listed below.
190
CHAP. 8
4
k3
51
0
k7
- k3
k1+k2+k3
k1
k1
k2
k1+k4+k5
---k4
k4
k5
k2
k2+k4+k6 +k7+k8
k6
k6
k3
k5
k3+k5k6 +k9+k10
k9
k9
k10
k7
+k7+k9
+k11
k11
0
k8
k10
k11
+k8+k10
+k11
Note that i(
3.
is
where contains the external joint forces. We start with i?PN If joint q is un restrained, we put in row q If joint q is fully restrained, we modify and according to (815) and (816). Finally, if joint q is partially restrained, we use (819) through (821). Since is symmetrical, we have to list only the submatrices on and above the diagonal. It is convenient to work with successive joint numbers. For this system,
joint 2 is fully restrained and joints 4, 6 are partially restrained. The basic matrices for
joints 4,6 and the initial and final forms of.Yt', are listed below. Note that this procedure does not destroy the banding of the stiffness matrix.
Joint 4
(u42 is prescribed)
R4
E4=[
ri
01
oj
6
G4=[0
=
[0
= {O,ii42j
Joint
1
(t42 is prescribed)
1
[
01
ii ii
=
ri E6=[0
oJ
SEC. 84.
INCREMENTAL FORMULATION
ir22
)r33
ir34
.X44
ir4,5
Sym
(U4)
(Us)
ir,4E4
U
1,0
.K13
.K34E4
J35
E4ir46(E6R6)T + G4 +
--
ir56(E6R6)T
(E6R6)iq66(E6R6)T
+ G6
E6R6(
84.
Equations (813), (822) are valid for both linear and nonlinear behavior. However, it is more efficient with respect to computational effort to employ an incremental formulation when the system is nonlinear. With an incremental formulation, one applies the load in increments and determines the corresponding displacements. The total displacement is obtained by summing the displacement increments. An incremental loading procedure can
also be used with (813) but, in this case, one is working with total displacement
rather than with incremental displacement. In this section, we develop a set of equations relating the external load and the resulting incremental displacements. These equations are also nonlinear, but if one works with small
192
CHAP. S
the incremental system equations. We complete the section with a discussion of the classical stability criterion. We start with (84), which defines the external joint forces required to equilibrate the action of the force for bar n,
generate
Pn*
p,,
displacement for the new equilibrium position. Since F and depend on U, their values will change. Letting AF, AD be the total increments in F, D due to AU, and requiring (a) to be satisfied at both positions, leads to the following incremental force-equilibrium equations:
=
Ap,,.. =
Afif +
+
D.
(824)
The exact
=
ci,,
u,,_)
To allow for the possibility of retaining only certain nonlinear terms, we write (a) as
fi,,
= = =
(u,,
u)Tg,
u,,..) +
u,,)
(825)
Yh(ufl. u,,_)
To neglect a particular displacement component, we delete the corresponding element in For geometrically linear behavior, = 0. Operating on (825),
we obtain
dv,, =
Au,,)Tg,,
(826)
and
Ae,, =
=
d2e,,
Au,,)
(827)
It remains to evaluate AF,,. We allow for a piecewise linear material and employ the relationst developed in Sec. 64. For convenience, we drop all the
t See (631), (632), and (633).
SEC. 84.
INCREMENTAL FORMULATION
193
where k, are constant for a segment. They have to be changed if the limit is unknown, one of the segment is exceeded or the bar is unloading. Since has to iterate, taking the values of k, Ae0 corresponding to the initial equilibrium position as the first estimate. This is equivalent to using the tangent stiffness. The initial elongation, Ae0, is included to allow for an incremental temperature change. Substituting for (828) takes the form
At' _AC'
Finally, we substitute for
LXPn+ =
+ if
72
Ct
(829)
Lw,, in
&i,..) +
(830)
where
= Fag,, + =
(831)
-i-
We interpret k7 as the tangent stiffness matrix. The vector, L\p9, contains linear, quadratic, and cubic terms in We have included the subscript g to indicate that it is a nonlinear geometric term. We write the total set of incremental joint equilibrium equations as
=
where
+ M'0 +
(8-32)
is assembled using (89) and MPO + with (88). Note that is symmetrical. Finally, we introduce the displacement restraints by ap-
(8-33)
It is convenient to include the prescribed incremental support displacement terms in involves only the incremental temperature and so that the variable displacement increments. The contracted equations are
K1,11 AU1 =
where K1,
AP0,1
AP9,1
1(1,12
AU2
(8-34)
since contains quadratic and cubic terms in MI. There are a number of techniques for solving nonlinear
194
CHAP. 8
which is the easiest to implement, but its convergence rate is slower in com-
parison to most of the other methods. First, we note that and are independent of A'1/1. They depend only on the initial equilibrium position and the incremental loading. We combine and and write (833) as
X7K MI1 =
(8-35)
Now, we let represent the nth estimate for LXa/IJ and determine the (n + 1)th estimate by solving
)p* L\cW(n+
&?P
(836)
The iteration involves only evaluation of and back-substitution once is transformed to a triangular matrix. The factor method is particularly convenient since X7 is symmetrical. With this method,
STS
(837)
= STQ = A9*
S
(838)
(8-39)
as the "actual" displacement increment. One can interpret this scheme as one
cycle successive substitution. The solution degenerates when the tangent stiffness matrix becomes singular.
To investigate the behavior in the neighborhood of this point, we apply the classical stability criterion developed in Sec. 76. The appropriate form for a
truss is given by (741):
den)> 0
0 (a)
We have already evaluated the above terms. Using (826), (827), and (829) with Ae0 = 0,
+
and (a) can be written as
d2WD
ALT ICE,
Au,,)
(b)
AU1 > 0
(840)
It follows that
Iterative techniques are discussed in greater detail in Secs. 187, 188, 189.
_____
SEC. 84.
iNCREMENTAL FORMULATION
But K1,
and
where H is a nonsingular permutation matrix, which rearranges the elements of according to (8-42) =H
Then, and K1 have the same definiteness Finally, we can classify the stability of an equilibrium position in terms of the determinant of the tangent stiffness matrix:
D
D>O
D=0 D< 0
85
= 1K1,
iii
(843)
Example
We illustrate the application of both the total (813) and incremental (834) formulations
to the truss shown in Fig. E85A. To simplify the analysis, we suppose the material is linearly elastic, k1 = k2 = k, and there are no initial elongations or support movement.
b <<d
Fig. E85A
x2
d_____
b]
+ =
= cc, +
fl = 1,2
for this example. Since b cc d, we can neglect the nonlinear terms due to u11, i.e., we can take 1 [0 0
1
196
CHAP. 8
Using (c),
1
I
Pi
+ u12]
12
112
b +
=
= +
= ky,,u1
n = 1,
Finally, the force-equilibrium equation for joint I follows by applying (86) to both bars.
(k1
Jusi
0
u12)(b
and
F1 =
(h
F2 =
(b
The diagonal form of the coefficient matrix is due to the fact that we neglected u11 in the expressions for y and This approximation uncouples the equations. Note that (g) is the first equation in (813) with U2 and P0 set to 0. Solving the first equationt in (g). we obtain
I (J\2
The corresponding bar forces are
Pti
Pi
=
This result is actually the solution for the linear geometric case. The expression for and the corresponding bar forces follow from the second equation
[2k
P12
u12)(b
1u12)j 012
0.
SEC. 54.
F1
INCREMENTAL FORMULATION
197
P12
F2
(b
2u12)u12
2bu12
L
= 2k (b
u12)(b
and solve (m) by iteration. Alternatively, one can specify u12 and evaluate from (k). The latter approach works only when there is one variable. The solution is plotted in Fig. E8-5B.
P12
A
9
Fig. E85B
(I +
0
b
We describe next the generation of the incremental equations which follow from (826) (832). Applying (826), (827) to (b)(d) results in
= d[i,, =
de,
= [o
a,1 Au11
=
Au1
d2e, =
(Au1 )2
n=l,2
We arc assuming no initial elongation. Then,
k
k(de,,
The tangent stiffness matrix and incremental geometric load term are defined by (831). Using (n), we obtain
+ u12)
= Sym
(Au12)2
+ u12)2 +
198
CHAP. 8
Finally, we assemble the incremental equilibrium equations for joint 1 using (830).
Ap1 =
+ k, 2)Au1 + Ap1, i +
Al)9. 2
(r)
?4(_b + u12)2 +
+ F2)
0
Note that (s) is (834). Also, the incremental equations are uncoupled. We restrict the analysis to only 112 loading. Setting F1 = F2 in (s) results in
+ u17)2)] Au12 =
Ap12
[Au12 + 3(b +
(t)
where F is determined from (e). The coefficient of Au12 is the tangent stiffness with respect
to u12.
du12
+ u12)2)
(u)
>
du12
<
(v)
Points A, B are stability transition points and the segment A-B is unstable. If k = 0, the truss is neutral with respect to Au11. Now there is a discontinuity in k at F = Feb, the pin-ended Euler load, when the material is linearly elastic:
k=0
,t2EI3
(w)
F4 =
AE(
\2
AEb2
)=
=
(x)
Then, for system instability rather than member instability to occur, b must satisfy
b
(y)
SEC. 84.
INCREMENTAL FORMULATION
199
Lastly, we outline how one applies the method of successive substitution to (t). For convenience, we drop the subscripts and write (t) as
Ltu =
0.
=
The second estimate is determined from
(aa)
=
Generalizing (bb),
(bb)
The convergence is illustrated in Fig. E85C. Case (b) shows how the scheme diverges
Fig. ES5C
(a)
Lip
Ib)
in the vicinity of a neutral point = 0). Convergence generally degenerates as and one has to resort to an alternate method.
200
CHAR 8
85.
In the previous section, we illustrated the behavior of a geometrically nonlinear system. The analysis involves first solving the nonlinear equilibrium equations for the displacements and then applying the classical stability criterion to determine the stability of a particular equilibrium position. Once the nonlinear equilibrium equations are solved, the stability can be readily determined. Now, if a geometrically nonlinear system is loaded in such a way that it behaves as if it were geometrically linear, we can neglect the displacement that is, we can take = in the expression for terms in This approximation is quite convenient since we have only to solve the linear problem in order to apply the stability criterion. We refer to this procedure as linearized
stability analysis.
+ F,,g,,
(844)
The first term is the linear stiffness matrix. We interpret the second term as a geometric stiffness. The bar forces are determined from a linear analysis of the truss. If the loading is defined in terms of a single load parameter, we
can write (844) as
F,, = k,,,, =
(845) k1,,,
+ ).k9,
The tangent stiffness matrix is generated by applying the Direct Stiffness Method to each term in (845). We express the actual and modified matrices as
K,,
= =
K1,
,i + Kg, ii
+
(846)
and
(847)
where K1 is the system stiffness matrix for linear behavior. It is symmetrical and positive definite when the system is initially stable. The geometrical stiffness, K9, is also symmetrical but it may not be positive definite. Equation (846) shows that the tangent stiffness matrix varies linearly with the load parameter. If the system is initially stable, K,, is positive definite for 2 = 0. As 2 is increased, a transition from stable to neutral equilibrium may occur at some load level, say 2cr To determine we note that neutral equilibrium (see (843)) corresponds to K,, = 0 which, in turn, can be interpreted as the existence of a 'non-trivial solution of
K,,
=0
(a)
SEC. 85.
201
AU1 =
and
Since 1K1, itt
).K5,11 AU1
(848)
Mt'
1,1
Ault' =
Mlii
I
(849)
0 1JAU51
U
[o
Premultiplying by H =
AHT
[Kg, ii
01 5Au1
[o
i,j
l,,AU2
(a) becomes
(na eqs.)
(b)
(c)
AU2 =
The solution of (c) is
2 AU2
(r eqs.)
AU2C1
o
+C2
o
Example
86
Consider the system shown. We suppose the bars are identical, the material is linearly elastic, and there is no support movement, The geometry change is negligible under a vertical load and we can use the linearized stability criterion. Working with the undeformcd geometry, we have
F1 =
=
F2
Matrix iteration (Ref. 1) is a convenient computational scheme for determining 2,,. We apply
(K9,,1)AU1 =
AU1
jtto
202
CHAP. &
Fig. 8B
d <<b
12
We let k1 = k2 = k. The system stiffness matrices follow from (844) and (845).
=
21
+ k 2 =
0
+
(b)
and
k8,1
+ k9,2 =
(g5
+ g2)
(c)
+ ufg,,
(d)
We neglect u12 in the general expression for approximate expressions for fi, and g, arc
L
b]
b]
(e)
i[l
0 0
0
(d)2
= 11e,ii + AKj,ji =
2k
0
(b)2
(g)
SEC. 85.
203
is semidefinite) occurs at
2kb
\\L)
\,LJ
Note that (g) has only one eigenvalue instead of two. This is a consequence of our using approximate expressions for Equations (e) instead of the exact expressions. At 2 = the system is neutral with respect to Au1 i.e., the buckling mode is antisymmetric. Neutral equilibrium also occurs when the bars either buckle or yield. The value of 2 for Euler buckling of the bars is
eb
= L
2b
2b k2EIl
2AEb/irp'\2
LLL2J
I
L \L
Comparing (h) and (i), we see that Euler buckling of the bars controls when
d> irp
The exact expression for g, is
I,.
(d)2
K1,
= K,,
2k
(b)2
In this case, there are two characteristic values and therefore two critical values of 2.
1
= 2kb
2kb
(d\2
(Ii)
2cr. 2
C)
Acri
The second root corresponds to neutral equilibrium with respect to Au12. For this example, b, and the first root defines the critical load. It is of interest to compare 2cr 2 with the buckling load found in Example 85. There we considered d b and followed the nonlinear behavior up to the point at which the slope of the P12 u12 curve vanished (neutral with respect to Au12):
d
1(1
du12
0
mex
(0)
The linearized result is significantly higher than the true buckling load. In general, the linear buckling load is an upper bound. How close it is to the actual value will depend on the geometry and loading. When d << b, it is quite close, while it considerably overestimates the true load for d b.
204
CHAP. 8
REFERENCES
C. FL, and J. B. WILCUR Elementary Structural Analysis, McGraw-Hill, New York, 1960. 2. HALL, A. S.. and R. W. W00DHEAD: Frame Analysis. Wiley, New York, 1967, 3. ARGYRIS, J. H. and S. Kelsey: Energy Theorems and StruciuralAnalysis, Butterworths, London, 1960. 4. LIVESLEY, R. K.: Matrix Methods of Structural Analysis, Pergamon Press, 1964. 5. DC VEUBEKE, B. F., Matrix Methods of Structural Analysis, Pergamon Press, 1964. 6. MARTIN, H. C. Introduction to Matris Methods of StructuralAnalysis, McGraw-Hill, New York, 1965. 7. ARGYSIS, J. H.: Recent Advances in Matrix Methods of Structural Analysis, Pergamon Press, 1964. 8. RUBINSTEIN, M. F.: Matric Computer Analvsis of Structures, Prentice-Hall, 1966. 9, PRZEMIENIECKI, J. S.: Theory of Matrix Structural Aizah'sis, McGraw-Hill, 1968. 10. THOMPSON, J. M. T. and A. C. WALKER: "The Nonlinear Perturbation Analysis of Discrete Structural Systems," Jut. J. Solids Structures, Vol. 4, 1968, pp. 757768. 11. RUBINSTEIN, M. F.: Structural ,S'vsse,'nsStatics. Dynamics, and Stability, PrenticeHall, 1970. 12. RALSTON, A.: A First Course in Numerical Analysis, McGraw-Hill, 1965.
1.
PROBLEMS
81.
linear.
(a)
Express
= Vr
VT
PTU1
e0,1)2 =
eo)Tk(e
e0)
(b)
Note that
de =
BfAT.11
(c)
82.
Determine and F due to a temperature increase of 100F for all the bars. Assume no support movements at joints 2, 3, 4. 83. For the structure sketched: Determine the displacements, bar forces, and reactions.
84.
(a) (b)
\EaJ For the structure sketched: Develop the general form of Indicate how you would obtain K11.
PROBLEMS
205
Prob. 82
E=3X
Bar areas = 3
ksi
X 106/F
Prob.
20'
10'
83
E=3X
ksi
Initial elongation of
in.
to the left
I
0
6 kips
x2
Prob. 84
807
605
206
CHAP. 8
20'
Prob. 85
Area
3a
I
Y3
2 3
4a
3a
4
5
4a
2.5a
Prob. 8-6
87. Determine the load-deflection relation for the system shown. Consider the material to be linearly elastic and the bars to bc identical, Assume no initial elongation or support movement. 88. Investigate the elastic stability of the system shown. Assume the material is linearly elastic and no support movements. Use the linearized stability criterion and work with the exact expression for Rework the
problem, considering d b and using the corresponding approximate expression for 89. Determine the lowest critical load for the truss shown. Assume the material is linearly elastic and all bars have the same stiffness.
wp
Prob.87
d <<b
k1 Ic2
_AE
-y-
x1
K2
2
Prob. 8S
Prob. 89
T
20'
208
CHAP. 8
810. The governing equations for geometrically nonlinear behavior of a linearly elastic discrete system such as a truss are nonlinear algebraic equations containing up to third-degree displacement terms. We have expressed them as
contain linear displacement terms. This form is dictated by our choice of matrix notation. In order to expand (a), we must shift from matrix to indicial notation. For convenience, we employ the summation convention. If a subscript is repeated in a term, it is understood the term is summed over the range of the repeated subscript. An example is
where d,
(j = 1,
n)
We write the ith equilibrium equation for tile system as (this representation is suggested in Ref. 8-40):
+ KIJkU,, +
where i, j,
2P1
k, range over the total number of unknowns, U1 is the total value of the jth displacement unknown, 2 is a load parameter, defines the load distribution, and the K's are constants which can be interpreted as second-, third, and fourth-order tensors. The second-order tensor, is the linear
stiffness matrix.
(a)
We generate the system tensors by superimposing the contribution of each bar. The first step involves converting the matrix expressions
p,,
=
k,,e,, + F0,,,
'y,,(u,,
p,, - = p,,
(d)
where
F,,
e,, =
u)T
l.Pn_j
(U,,J
(f)
(k,1
+ Po,
PROBLEMS
209
Show that
k11 =
=L
=
where c is defined by
[1
(Ii)
{::} = cu
(i)
Discuss how you would locate the appropriate addresses for the bar stiffness tensors in the system tensors. What symmetry properties do the k's exhibit? Do these properties also apply for the system tensors? (b) Develop the incremental equations relating Au, AA and compare with
(830).
(c)
811.
Specialize the incremental equations for linearized stability analysis. For the structure sketched:
Prob. 811
=k
(a)
(b)
Determine the nonlinear incremental equilibrium equations at the equi0, P2 P2, Cr, the linearized critical load. Take Ap1 = 0 and solve for Ap2 as a function of Au1. Comment on how the system behaves when a small horizontal load, Pi ep2, is applied in addition to
GENERAL
The basic equations for the linear geometric case have the form
P1 =
B1F
e = BfU1 +
P2 = B2F
e0 + fF
where the elements of B1 and B2 are constants. Equation (a) represents linear equations relating the na prescribed joint forces and the in unknown bar forces.
that is, the rows of B1 must For the system to be initially stable, r(B1) = be linearly independent. This requires in In what follows, we consider the system is said to be statically determinate only stable systems. If in = since one can find the bar forces and reactions using only the equations of statics. The defect of (a) is equal to in nd = q, and is called the degree of
indeterminacy. One can solve (a) for na bar forces in terms of the applied forces and q bar forces. We refer to the system defined by the na bars as the primary structure and the q unknown forces as force redundants. In order to determine F, q additional equations relating the bar forces are required. These equations are called compatibility conditions and are obtained by operating on (b) which
The general procedure outlined above is called the Jbrce or flexibility method. This procedure is applicable only when the geometry is linear. In what follows, we first develop the governing equations for the force method
by operating on (a)(c). We then show how one can establish the compatibility equations using the principle of virtual forces and discuss the extremal character of the force redundants. Finally, we compare the force method for a truss with the mesh method for an electrical network.
210
SEC. 92.
211
9-2.
We consider the first columns of B1 to be linearly independent (if the system is initially stable, one can always renumber the bars such that this
=[B11
ii
B12]
I
"a)
(na x q)
B2
(rxm)
[B21
(rxn4)
(na x 1)
(rxq)
B22]
=
F2
(qx 1)
The bars corresponding to F1 comprise the primary structure and F2 contains the q redundant bar forces. Using (91), the force-equilibrium equations ((a) and (c)) take the form
B11F1 =
P1 B12F2
(ad eqs.)
(92)
(93)
P2 =
B21F1
+ B22F2
(reqs.)
Since Bit! 0, we can solve (92) for F1, considering P1, B12 as righthand sides. The complete set of q + 1 solutions is written as
F1 = F1,0 + Fl,F,F2
where F1.0 and F1, F2 satisfy
(94)
Note that the kth column of F1, F2 contains the bar forces in the primary structure due to a unit value of the kth element in F2. Also, F1, contains
the bar forces in the primary structure due to the applied joint loads, P1, with
F2 =
0.
P2 = P2,o =
P2,0
+
(96)
B21F1,0
= B21FI,F2 + B22
We consider next (b). Partitioning e, e0, and f,
(nixl)
e
(fl4X 1)
e1
C2
j (q< 1)
ii
(97)
L 0
f2
(q x q)
212
CHAP. 9
(ne, eqs.)
(9-8)
(99)
(q eqs.)
Once e1 is known, (98) can be solved for U1. We obtain the equation for F2 by eliminating U1 in (99). First (see (95))
we note that
BT 12
12
1,F21 IF2
\T_
T 11
P2, F2U2 =
ez
+ Ff
+ f2F2 + Ff,FZ(CLO + f1F1)
e2,0
(911)
The first form, (910), shows that the equations are actually restrictions on the elongations. One can interpret (910) as a compatibility condition, i.e., it must be satisfied in order for the bars tofit in the deJbrmed structure defined by U1. The second form, (911), follows when we express the elongations in terms of the bar forces. Finally, we substitute for F1 and write the result as
122F2
d2
C
(912)
T
where
f 122
.1
t'T
U2 e2,0
+ nT
f'r
1O...43
The coefficient matrix, f22, is called the flexibility matrix for F2. One can show that f22 is positive definite when the bar flexibility factors in f2 are all positive.t If the material is physically nonlinear, f. and e0,, depend on F1. Iteration
is minimized by applying the loading in increments and approximating the force-elongation relation with a piecewise linear representation. The incremental equations are similar in form to the total equations4 We just have to replace the force, displacement, and elongation terms with theit incremental values and interpret las a segmental (tangent) flexibility. At this point, we summarize the steps involved in the force method.
I.
Determination of F1, 0, P2. 0, F1.
and P2, F2
We select a stable primary structure F1 and determine the bar forces and
reactions due to P1 and a unit value of each force redundant. This step involves q + 1 force analyses on the primary structure. Note that we obtain the primary
structure by deleting q = ni bars. The selection of a primary structure and solution of the force equilibrium equations can be completely
t See Prob. 9i.
We reduce
X See Prob. 94.
SEC. 92.
GOVERNfNG
APPROACH
213
2.
We assemble d2, and solve f,,F, = d, for F,. Then, we determine F, and P2 by combining the q + basic solutions.
F1
P2
3.
Determination of U,
e1,0 +
f1F1
e,
for U,. If only a limited number of displacement components are desired, one can
determine these components without actually solving (98). To show this, we write U, as
U1't
e1
21
hT 111
We see from (915) that the kth column of Br,' contains the bar forces in the primary structure due to a unit value of the kth element in P1. Also, it follows from (96) that the kth column of B2 contains the reactions due to a unit value of the kth clement in P,. Now, we obtain the kth element in U1 (which corresponds to the kth element in P,) by multiplying the kth column of Br1' by the kth column of B21Br,' by Of. and adding the two scalars. Then, letting
F,pJk = F,
P2.pj,, =
P2
(914)
nT
fl
Note
Step 1:
in
F1={F,,F2}
F2={F3}
The primary structure consists of bars I and 2. Note that all force analyses are performed on the primary structure. The forces and reactions corresponding to P, and F3 = + 1 can be readily obtained using the method ofjoints. The results are shown in Fig. E91B.
214
CHAP. 9
Fig. E91A
10 kips
20 kips
(1) A1 = 1.Oin,2
A2 0.51n.2 A3 =0.5 in.2 (2) Material is linearly elastic. E = 3 )< i04 ksi for all bars. = 0. (3) e0,1 1/16 in. eo,2
1/2
2.50
3/8
20.83
10 kips
20 kips
B12F'2
110)
[-.8 +.8flFj =
Step 2:
+ [1j{F3}
[01
Since only u32 and u41 are finite, we can contract U2 and P2,
=
u41} = {u3,.
and write
PT
fl
/D'
2
215
{2083, 4.17}
{
(kips)
(kips) (kips)
=
Also, we are given that
e1, o
{eo,
e0,
2} =
T'6, 0}
(inches) (inches)
e2,0 = {eo,3} = 0
f2 and evaluate f22 and d2. It remains to assemble The flexibility factors are (in./kip)
12(25)
12(25)
=
Then,
12 =
13
12(20)
=
f2
[f3] =
0.8(2 x
F2 = {F3} = 5.27kips
F1 = F1,0 +
1 17.53kips
0.S7kips
Equation (a) actually represents a restriction on the elongations. The original form of (a)
follows from (910).
8e1 a
Equation (b) reduces to (a) when we substitute for the elongations in terms of the bar forces.
= Ff, p,1e1
Now,
)T
(c)
fr'2 (1
UO'
15
= e1,0 +
f1F1
{-.-.24, .018}
We apply a unit load at joint 1 in the X1 direction and determine the bar forces in the
primary structure and the reactions (P32, P41) corresponding to the nonvanishing prescribed displacements
1
216
CHAP. 9
a11 = +185
to solving (98).
.033
= +15 in
If both displacement components are desired, we apply (915) twice. This is equivalent
93.
We obtained the elongation compatibility equations (910) by operating on the elongation-displacement equations. Alternatively, one can use the principle of virtual forces developed in Sec. 73. It is shown there (see Equation (714)) that the true elongations satisfy the condition,
AFTe
for any statically permissible system of virflial bar forces and reactions which satisfy the constraint condition, B1 AF =
AP1
Equation (b) states that the virtual bar forces cannot lead to increments in the prescribed jOint loads, i.e., they must be self-equilibrating.
Now, using (94), (95), we can write
F
where
{Fio}
+
B1
and
B1
Then
AF2
= satisfies (b) for arbitrary AF2. The reactions due to AF2 are obtained from
(96):
A?2 =
B2
AF =
P2
F2
Equation (h) must be satisfied for arbitrary AF2. Finally, it follows that
FT,F,el + C2
P2 F2U2 =
(i)
SEC. 94.
217
linear, (i) leads to a set of q linear equations in F2 when we substitute for the elongations in terms of the bar forces. We determine the displacements by applying the general form of the principle
of virtual forces (see (710))
AFTe
= APfU1 =
=
AF2 = 0
and (j) leads to
U1 =
Note that
I
1)11
One can interpret the compatibility equations expressed in terms of F2 as the Euler cquations for the total complementary energy function,
Pf02 =
= P2,F, = and (737) coincides with (i). We have written the expanded form of (i) as
f22F2 =
Since (i) are the Euler equations for
d2
=
and it follows that
d2)
2U2
H _I T 221 2 2
c
for the linearly elastic case. One can show that the stationary point corresponds to a relative minimum value of H. when the tangent flexibility factors for the redundant bars are all positive. t
It is of interest to compare the force method for a truss with the procedure followed to find the currents in an electrical network. The latter involves the
t See Prob. 98.
218
CHAP. 9
application of Kirchhoff's laws and is called the mesh method. Various phases of the electrical network formulation are discussed in Probs. 66, 614, and the
governing equations for a linear resistance d-c network are developed in Probs. 614, 623. We list the notation and governing equations for convenience:
b = number of branches
N=n1 MbN=bn+1
k+, k
number of nodes
= potential at node j with respect to the reference potential, = nodes at positive and negative ends of branch k = current in branch k, positive when directed from node k_ to node = potential drop for branch k =
Vk.
Vk+
ek
e0,
The governing equations expressed in matrix notation are (see Prob. 623):
An = 0
(N eqs.)
e0
(916) (917)
e = AV =
where
i
+ Ri
(b
eqs.)
e = {ej, e2,
(918)
v={V1,V2
R1
R=
R2
Rb
and A is obtained by deleting the last column of the branch-node connectivity matrix d. Note that d has only two entries1 in any row. For row k
(Ic =
jlkk = +1
dkk+ =
dkj
1 1
k+ ork.
(919)
=0
1= 1,2,...,N
Actually, d is just the matrix equivalent of the branch-node connectivity table. Example
92
A network can be represented by a line drawing consisting of curves interconnected at various points. The curves and intersection points are conventionally called branches and nodes respectively. Each branch is terminated at two different nodes and no two branches have a point in common which is not a node. Also, two nodes are connected by at least one path. A collection of nodes and branches satisfying the above restrictions is called a linear connected graph, If each branch is assigned a direction, the graph is said to be
SEC. 94.
219
oriented. The connectivity relations for a network are topological properties of the corresponding oriented graph. Consider the oriented graph shown. We list the branch numbers vertically and the node numbers horizontally. We assemble d working with successive branches. Finally, we obtain A by deleting the last column (cot. 4) of d.
2
Fig. E92
Node
Branch
1
+1
2
3
+1 +1
.1
4
1
(bxN)
[ A1
[A2 (MXN)
(Nx 1)
= JJi__
(Mx))
220
CHAP. 9
ATi1 =
(921)
0, we can solve for i1 in terms of i2. We write the solution of the Since tAil node equations as
i=
=
Ci2
[Cii. I'2
I
(922)
C1 =
(923)
It remains to determine a set of M equations for i2. One can express (917) in partitioned form and then eliminate V, or alternatively, one can use the variational principle developed in Prob. 76. Using the first approach, we write (917) as
= A1V =
e2
e1
+ R1i1
(N eqs) (M eqs)
(924)
e1 =
e2 +
e1
+ R1i1
(925)
Eliminating V from the second equation in (924) and using (923), we obtain
CTe1
(926)
Equation (926) represents M equations relating the branch potential differences (voltages). Finally, substituting for in terms of leads to
(R2 + CTR1C1)i2
e2,0 Cfe10
(927)
The coefficient matrix for i2 is positive definite when the branch resistances are positive. This will be the case for a real system. The essential step in the solution involves solving (921), that is, finding C1. Note that C1 corresponds to for the truss problem. Also, the branches comprising A1 (and i1) correspond to the primary structure. Although the equations for the truss and electrical network are similar in form, it should be noted that the network problem is one dimensional whereas the truss problem involves the geometry as well as the cpnnectivity of the system. One can assemble C1 using only the topological properties of the oriented graph which represents the network. To find the corresponding matrices (F1, 0 and for a truss, one must solve a system of linear equations. In what follows, we describe a procedure for assembling C1 directly from the oriented graph. A closed path containing only one repeated node that begins and ends at that node is called a mesh. One can represent a mesh by listing sequentially
SEC. 94.
221
Let hT be the number of branches in a tree connecting 11 nodes. One can easily show that
meshes.
bT=nl=N
(928)
We reduce a graph to a tree by removing a sufficient number of branches such that no meshes remain. The branches removed are generally called chords. The required number of chords is equal to b = b N = M. Now, we associate the branches comprising a tree with the rows of A1. Selecting a tree is equivalent to selecting N linearly independent rows in A. The M chords correspond to the redundant branches, that is, the rows of A2. Note that one can always number the branches such that the first N branches define a tree. Chord j and the unique path (in the tree) connecting the terminals of chord j define a mesh, say mesh j. We take the positive direction of mesh j (clockwise or counterclockwise) such that the mesh direction coincides with the positive direction for chord j. Now, the current is constwit in a niesh. Suppose branch r is contained in mesh j. Then, the current in branch r due to a unit value of is equal to + 1 (1) if the positive directions of branch rand meshj coincide (are opposite in sense). We have expressed the solution of the node equations as
i1=C1
I)
(NXM)
(Mxl)
12
Now, we take the elements of i2 as the chord (mesh) currents. Then i1 represents
the required branch currents in the tree. We assemble C1 working with the columns. The column corresponding to involves only those branches of the tree which are contained in mesh j. We enter (+1, 1,0) in row k of this column if branch k is (positively, negatively, not) included in mesh j.
Example 93
For the graph in example 92, N = n 1 = 3 and b = 6. Then M = b N = 3 and we must remove 3 branches to obtain a tree. We take branthes 4 , 5 , and 6 as the chords. The resulting tree is shown in Fig. E93. We indicate the chords by dashed lines.
2
Fig. E93
222
CHAP. 9
ij =
i2
The meshes associated with the chords follow directly from the sketch:
mesh4 mesh5
mesh6
To assemble C1 we list the branches of the tree vertically and the chord numbers horizontally. We work with successive columns, that is, successive chords. The resulting matrix is listed below. Note that C1 is just the matrix equivalent of(a).
Branches
2
3
+1
0
1
ofthe
tree
1
+1
0
+1
1
A1=
0
1
+1 +1 +1
0 0 0
0
0
l
0
1
1
A2=
One can readily verify that
+1
C1 =
The matrix, C = {C1, Im}, is called thc branch-mesh incidence matrix. Using (923), we see that A and C have the property
(N x Ill)
ATC
(929)
CTe
(930)
The rows of CT define the incidence of the meshes on the branches. Equation (930) states that the sum of the potential drops around each mesh must be zero and is just Kirchhoff's voltage law expressed in matrix form. The matrix
PROBLEMS
223
that one would obtain by applying Kirchhoff's current and voltage laws to the various nodes and meshes. This, of course, also applies to the truss problem. The two approaches differ only with respect to the assemblage of the governing equations. In the conventional approach, one assembles the equations individually. This involves repeated application of the basic laws. When the equations are expressed in matrix form, the steps reduce to a sequence of matrix multiplications.
REFERENCES
1.
2. 3.
4.
5. 6.
7.
C. H., and J. B. WILBUR: Elementa,'v Structural Analysis, McGraw-Hill, New York, 1960. HALL, A. S., and R. W. W000HEAD: Frame Analysis, Wiley. New York, 1967. MORICE, P. B.: Linear Structural Analysis, Ronald Press, New York, 1969. RUBINSTEIN, M. F.: Matrix computer Analyris of Structures, Prentice-Hall, 1966. PRZEMIENIECKI, J. S.: Theory of Ma/rLr Structural Analysis, McGraw-Hill, 1968. RUBINSTEIN, M. F.: Structural SystemsStatics, Dynamics, and Stability, PrenticeFlaIl, 1970.
Di MAGGIO, F. D., and W. R. SNLLERS: "Network Analysis of Structures," Eng. Mech. Div.. A.S,C.E., Vol. 91, No. EM 3, June 1965.
FF.NVE.S, S. .1., and F. H. BRANIN, JR.: "Network-Topological Formulation of Struc-
8.
tural Analysis," J. Structures Div., A.S.C.E. Vol. 89, No. ST4, August 1963.
PROBLEMS
91. Show that the coefficient matrix f22 is positive definite for arbitrary rank of F1 P2 when is positive definite. Use the approach suggested in
Problems 212 through 214. 92. Solve the following system using the procedure outlined in Sec. 92. TakeX1 {xi,x2}
2
2
3
0
0
l.Y25
0 0 0
x1
12
00 1
x2
X3
+
1
0002 x4
93. Consider a system of in equations in n unknowns, ax = c, where in> n. Suppose r(a) = a and the first a rows of a are linearly independent.
Let q =
(a) (b)
n.
224
CHAP. 9
94.
=
Ae
B2
AF AF
Bf AU1 +
AU2 =
+ ft AF
represent the flexibility factor and incremental initial elongation where f', for the segment corresponding to the initial value of F. One has to modify bothft and if the limit of the segment is exceeded (see sec. 64 for a detailed
treatment). Consider the case where the loading distribution is constant, i.e., where only the magnitude is increased. Let P1 1,/i where 2 is the load parameter and
defines
computational scheme. Also discuss how you would account for either yielding
or buckling of a bar. Distinguish between a redundant bar and a bar in the primary structure. 95. Solve Prob. 83 with the force method: Take F3 as the force redundant. 96. Assemble the equations for F2 = (F8, F9, F10, F, for the truss
shown.
Prob, 96
x2
xl
15'
15'
(1) Material is linear elastic and the flexibility factors are equal.
= {u42 } (2) Only u42 is finite. Take (3) Only initial elongation for bar 4.
For the truss shown: Using (910), determine the elongation-compatibility relations. Takt as the redundant bars. bars , (b) Express u52 in terms of the elongations and support movements.
97.
(a)
98.
= AFTe
PROBLEMS
225
Prob. 9i
x2
x1
Then
= AFT de
Express d2fl, as a quadratic form in AF2. Consider the material to be nonlinear elastic and establish criteria for the stationary point to be a relative minimum. 99. Consider the oriented linear graph shown.
Prob. 99
0
(a) (b) (c)
10
GENERAL
The formulation of the governing equations for the behavior of a deformable solid involves the following three steps:
1.
2.
Study of deformation. We analyze the change in shape of a differential volume element due to displacement of the body. The quantities required to specify the deformation (change in shape) are conventionally called strains. This step leads to a set of equations relating the strains and derivatives of the displacement components at a point. Note that the analysis of strain is purely a geometrical problem. Study of forces. We visualize the body to consist of a set of differential volume elements. The forces due to the interactions of adjacent volume elements are called internal forces. Also, the internal force per unit area acting on a differential area, say dAd, is defined as the stress vector,
this step, we analyze the state of stress at a point, that is, we investigate how the stress vector varies with orientation of the area element. We also apply the conditions of static equilibrium to the volume elements. This leads to a set of differential equations (called stress equilibrium equations) which must be satisfied at each point in the interior of the body and a set of algebraic equations (called stress boundary conditions) which must be satisfied at each point on the
surface of the body. Note that the study of forces is purely an equilibrium problem.
3.
Relate forces and displacements. In this step, we first relate the stress and strain components at a point. The form of these equations depends on the material behavior (linear elastic, nonlinear elastic, inelastic, etc.). Substitution of the strain-displacement relations in the stress-strain
relations leads to a set of equations relating the stress and derivatives of the displacement components. We refer to this
system as the stress-displacement relations.
229
230
CHAP. 10
The governing equations for a deformable solid consist of the stress equilib-
rium equations, stress-displacement relations, and the stress and displacement boundary conditions. In this chapter, we develop the governing equations for a linearly elastic solid following the steps outlined above. We also extend the variational principles developed in Chapter 7 for an ideal truss to a three-dimensional solid. In Chapter 11, we present St. Venarit's theory of torsion-fiexure of prismatic members and apply the theory to some simple cross sections. St. Venant's theory provides us with considerable insight as to the nature of the behavior and also as to how we can simplify the corresponding mathematical problem by introducing certain assumptions. The conventional engineering theory of
neering theory for an arbitrary planar member. Finally, in Chapter 15, we present the engineering theory for an arbitrary space member.
1O2.
a= b=
{Oj, 02
{h1,b2
101
aTb =
bTa
a1b1
+ a2h2 + ... +
To avoid having to write the summation sign, we introduce the convention that when an index is repeated in a term, it is understood the term is summed over the range of the index. According to this convention
(i = 1,2
and we write the scalar product as
aTb = a1b1
ii)
(102)
(103)
The summation convention allows us to represent operations on multidimensional arrays in compact form. It is particularly convenient for formulation, i.e., establishing the governing equations. We illustrate its application
below.
Example
1.
10i
c=ax
aisrnx n
SEC. 102.
231
=>
(b)
2.
frxTax
g = xrbx
The matrix form of the product, fg, is fg =(xTax)(xTbx) One could expand (d) but it is more convenient to utilize (b) and
write (c)
(c)
as
f=
g
bk(xkxe
fg =
=
3.
alJbk,xlxfxkx(
DIJk(XIXJXkX(
II
Using (b),
xT(aTa)x
H =
d=
and can be expressed as
c-c1 = 0f50
axxraT
= alkaf,xkx,.
=
According to the summation convention,
= AIJk(XkXe
=
Then, we can write (h) as
d11
+ d22 +
d11
= trace of d
AIIk(XkX(
H=
4. Let represent square second-order arrays. The inner product is defined as the sum of the products of corresponding elements:
Inner product
,) =
= +
+ g21e21 +
(m)
ejj over
to
one-dimensional
arrays.
Let represent a one-dimensional set of elements associated with an orthogonal reference frame having directions If the
232
CHAP. 10
corresponding set for a second reference frame is related to the first set by
k
=
we
cos
1, 2, 3
(104)
and it follows that the set of orthogonal components of a vector are a first-order cartesian tensor. We know that the magnitude of a vector is invariant. Then, the sum of the squares of the elements of a first-order tensor is invariant.
(106)
A second-order cartesian tensor is defined as a set of doubly subscripted elements which transform according to
=
An alternate form is
(107)
j. k.
1.2. 3
,n.
(108)
The transformation (108) is orthogonal and the trace, sum of the principal second-order minors, and the determinant are invariant.t
=
fl(2)
where
= =
b12
021
L
022
b22
032
7
b23
033
7
b11
b13
1733
SEC. 103.
233
Excluding rigid body motion, the displacement from the initial undeformed position will be small for a solid, and it is reasonable to take the initial Cartesian
Undeformed
dp
F' (Deformed)
i3
2
112
approach. Also, to simplify the derivation, we work with cartesian components for . Then,
ii
(1012)
Since we are in the deformed state is The corresponding line arid we can write following the Lagrange approach, p =
(1013)
The extensional strain, r, is defined as the relative change in length with respect
234
CHAP. 10
e(1 + 4c) = ap.kejk (1015) = One can readily establish that (eJk) is a second-order symmetrical Cartesian tensor4 direction and letting Taking the line element to be initially parallel to the represent the extensional strain, we see that
(no sum) = e0 (1016) -1) = To interpret the off-diagonal terms, we consider 2 initially orthogonal line elements represented by (see Fig. 102) and having direction cosines
d r',
xa
I
'/12
p.-
dp'1
x2
(1
are also called the See Prob. 104. It is known as Green's strain tensor. The elements, components of finite strain. They relate the difference between the square of the initial and deformed lengths of the line element, i.e., an alternate definition of Cjk
ds2 = 2eJkdxJdxk
SEC. 103.
235
We define as the angle between the lines in the deformed which is called the shearing strain, follows by state. The expression for taking the dot product of the deformed vectors.
COS
(it
J=
Y12 =
Substituting for
k)dsf
(1
(sum on k only)
=
(a)
(1 +
=
shows that
2e13
(1017)
is related to the
(1018)
(I +
Equations (1015) and (1017) are actually transformation laws for extensional and shearing strain. The state of strain is completely defined once the strain tensor is specified for a particular set of directions. To generalize these expressions, we consider two orthogonal frames defined by the unit vectors and (see Fig. 103), take the initial frame parallel to the global frame = ti), and let = 15 tk. With this notation:
+ (1 + +
(1019)
The strain measures (e, y) are small with respect to unity for engineering materials such as metals and concrete. For example, e for steel. Therefore, it is quite reasonable (aside from the fact that it simplifies the expressions) to assume r, y in the strain expressions. The relations for "small" strain are:
1
(1020)
=
Differentiating
+ ii
u,,Ji,,
with respect to
S
OP
+ Un, j)l,n
236
Cl-lAP. 10
and
4Um, ,u,,,, k
In order to simplify (1021), we must establish the geometrical significance of the various terms.
x3
t3
t2
/
+
t,3
X2
which define the rotation of the line toward the X2, K3 directions. The geometrical relations of interest to us are
012,
sin
0j3 =
033 1 +
1421
sine12
(1 + 81)cos 0j3
)2
(1 +
Also, by definition,
+ uj 1
i4,
e11
U1,
1+
i+
i+
and 03,
03,1
1
(1 +
013
SEC. 103.
237
= A=
1
(1
{1
A}112
(1023)
= I
(1024)
to (1
(1025)
In what follows, we assume small strain and express the derivatives and extensional strain (see Equation (d)) as
u3. 1 = 0(013)
U1
1=
0(012,
1
u1,
n2
t/12, "13
a11 =
(f)
dx1
dx1
1123 dx1
X1,u1
In the linear geometric case, the rotations are neglected with respect to strain. Formally, one sets 012 = 613 = 0 in (f) and the result is a linear relation between
(g)
238
CHAP. 10
with the initial orientation. The general relations for the linear geometric case
= =
as strain.
en
= =
(no sum)
+ ui,,
(1026)
0
1
(1027)
We can neglect
and
u11 +
(h)
The complete set of strain-displacement relations for small strain and smallfinite rotation are listed below for reference.
= = =
+ +
I
+
+ Uk iUk. j
k
(no sum)
(1028)
for a member in Chapter 18. Lastly, if no restrictions are imposed on the magnitude of the rotations, one must use (1021). The relations for finite rotation and small strain are = =
=
=
(no sum)
+ u11(l +
k
+ Uk.iUk,j
1029
Note that the truss formulation presented in Chapter 6 allows for arbitrary magnitude of the rotations. We have shown that linear strain-displacement relations are based on the
following restrictions:
1.
2.
The strains are negligible with respect to unity, and Products of the rotations are negligible with respect to the strains.
The first condition will always be satisfied for engineering materials such as metals, concrete, etc. Whether the second restriction is satisfied depends on the configuration of the body and the applied loading. If the body is massive in all three directions, the rotations are negligible with respect to the strains for an arbitrary loading. We have to include the nonlinear rotation terms in the strain displacement relations only if the body is thin (e.g., a thin plate or slender member) and the applied loading results in a significant change in the geometry. As an illustration, consider the simply supported member shown
SEC. 103.
239
in Fig. 105. We can neglect the change in geometry.if only a transverse loading
is applied (case 1). However, if both axial and transverse loads are applied (case 2), the change in geometry is no longer negligible and we must include the nonlinear rotation terms in the strain-displacement relations.
Case 2 (Q,P)
Case1 (Q)
geometry rather than the initial geometry. This can be defined by tracking
the movement of a triad of line elements initially parallel to the global directions. We let be the initial set and the deformed set (see Fig. 106). By definition,
= = =
we can write
vi
(no sum)
(no sum)
(1
+
Using (a),
(1030)
We will utilize (1030) in the next section to establish the stress equilibrium equations for the geometrically nonlinear case. Equations (1030) reduce to
(1031)
(no sum)
[032
240
CHAP. 10
/
ii
=dxji1
104.
ANALYSIS OF STRESS
The effects of the surroundings on a body such as contact pressure, gravitational attraction, etc.,result in internal forces. In this section, we establish the equilibrium conditions for the internal forces in a body. This step is generally called the analysis of stress. Consider a body subjected to some effect which results in internal forces. We pass a cutting plane through the deformed body and separate the two segments as shown in Fig. 107. We let in denote the outward normal direction for the internal face of body and refer to this face as the + in face. In general, the subscript, m, is used for quantities associated with the + m face. Now, we consider a differential area element AAm, and let A Fm be the resultant internal is defined as force vector acting on this element. The stress vector,
hm
AA, -. I)
(1033)
Note that has the units of force/area. Also, it depends on the orientation of the area element, i.e., on the direction of the outward normal. We do not allow for the possibility of the existence of a moment acting on a differential area element. One can include this effect by defining a vectort in addition to a stress vector.
f See Ref. 6, p. 68.
SEC. 104.
ANALYSIS OF STRESS
face.
241
We
From
Newton's law,
=
and it follows that
=
reversed.
,,,
(1O34)
The stress vector has the same magnitude and line of action but it's sense is
Body I
Body
In order to analyze the state of stress at a point, say Q, we need an expression for the stress vector associated with an arbitrary plane through Q. With this objective, we consider the tetrahedron shown in Fig; 108. The orientation of the arbitrary plane is defined by q, the outward normal direction. The outward normals for the other three faces are parallel to the reference axes (X1; j = 1, 2, 3). To simplify the notation, we use a subscript j for quantities associated with the X face, that is, the face whose outward normal points in the + direction. For example, we write
= =
(1035)
=
etc.
The force vectors acting at the centroids of the faces are shown in Fig. 108. The term M0 represents the change in due to translation from Q to the centroid. For equilibrium, the resultant force and moment vectors must vanish. In the
limit (as P
242
CHAP. 10
consider only the force equilibrium condition. From Fig. 108, we have
crq + &Tq
(a3 +
Now, AA1 is the projection of AAq on the X2-X3 plane. Noting that the projection of LxAq on a plane is equal to AAq times the scalar product of and the unit normal vector for the plane, and letting aqj be the direction cosine for the q direction with respect to the direction, we can write
AA.
cos(q,
X1) =
(1036)
(1037)
Once the stress vectors for three orthogonal planes at Q are known, we can determine the stress vector for an arbitrary plane through Q with (1037).
x2
+
i2
I
x3
2)LXA2
Equation (1037) is the transformation law for the Stress vector. The component of 5q in a particular direction is equal to the scalar product of 6q and a unit vector pointing in thedesired direction. Now, we express the stress vectors in terms of their components with respect to the coordinate axes (j = 1, 2, 3). j = 1,2,3 (1038) = aqklk
SEC. 104.
ANALYSIS OF STRESS
243
Note that the first subscript on a stress component always refers to the flice,
and the second to the direction. For example. a12 acts on the X1 face and
points in the X2 direction. The positive sense of the components for a negative The normal (a13) and in-plane (afk) comface is reversed since ponents are generally called normal and shearing stresses. This notation is
illustrated in Fig. 109.
x2
ta22
T
FIg. 109. Notation
cxqjajk
(1039)
Letting
and noting (1038), (a) expands to
We generalize (c) for two orthogonal frames specified by the unit vectors
(see Fig. 103) where
t. =
t3
-ii
cJ3k:k
(1040) face
Defining identifying
in the
direction and
(1041)
=
This result shows that the set
is a second-order cartesian tensor.
244
CHAP. 10
It remains to establish the equilibrium equations for a differential volume element. The equilibrium equations relate to the deformed state, i.e., we must consider a differential element on the deformed body. Since we have defined the stress components with respect to the global Cartesian directions, it is natural to work with a rectangular parallelepiped having sides parallel to the global
directions. This is shown in Fig. 1010. Point 0 is at the centroid of the element.
i3
dfl3
di73
dr13
ant
Fig. 1010. Differential volume element in Eulerian representation.
The stress vectors are considered to be functions of the deformedt coordinates We obtain the forces acting on the faces by expanding the stress vectors
about 0 and retaining only the first two terms4 Letting b denote the external force per unit volume and enforcing the equilibrium conditions leads to
(10-42)
and
(1043)
The scalar force equilibrium equations are obtained by expanding the vector
equations using (1038).
Force equilibrium
k= k=
1,
2, 3
(1044)
Moment equilibrium
23
(1045)
Moment equilibrium requires the shearing stress components to be symmetrical. Then, the stress tensor is symmetrical and there are only six independent stress measures for the three-dimensional case and three for the two-dimensional case.
t We arc following the Eulerian approach here. Later we will shift back to the Lagrange approach Second- and higher-order terms will vanish in the limit, i.e.. when the element is shrunk to a point.
SEC. 104.
ANALYSIS OF STRESS
245
Equations (1044) must be satisfied at each point in the interior of the body.
Also, at the boundary, the stress components must equilibrate the applied
surface forces. We define and write
as
Ph =
(1048)
Pnj
i3nk0kj
f=
1,
2, 3
(1048) represent boundary conditions is a reaction. on the stress components. If is prescribed, Our derivation of strain-displacement relations employed the Lagrange approach, i.e., we considered the displacements (and strains) to be functions of the initial coordinates (xi). The analysis of stress described above is based on the Eulrian approach, where the deformed coordinates are taken as the
independent variables. This poses a problem since the strain and stress measures are referred to different volume elements. Figure 10Il shows the initial and
Initia'
/
(I + 2)dx2
(1 +e1)dxj
4
Deformed
dX2[
dx1
U12
Lagrange
(I Eu'er
drj1
(I
x2
Fig. 1011. Comparison of Eulerian and Lagrangian representations for a volume
element.
f See Prob. 1012.
246
CHAP. 10
with the Lagrange strains, we must work with a nonorthogonal parallelepiped whose sides are parallel to the deformed line elements in the analysis of stress. Conversely, to be consistent with the Eulerian stresses, we have to refer the strain measures to nonorthogonal directions in the initial state. In the linear geometric case, we assume small strain and neglect the change in orientation due to rotation. The two approaches coalesce and we just have where is the direction cosine for the with to replace with and initial direction of the exterior normal. The linear equilibrium equations are:
+ hk
0
(1049)
CXj
For the geometrically nonlinear case, we work with stress measures referred to the deformed directions (see Fig. 106) defined by the unit vectors, We
define
as the stress vector per unit initial area acting on the face which
initially is normal to the direction, b* as the force per unit initial volume, and as the force per unit initial surface area. Figure .1012 shows this notation for the two-dimensional case. The stress and force vectors are considered to be functions of the initial coordinates (xe). The equilibrium equations at an interior point are + b* =
x
0
(1050)
(X1
(1
=
The set,
and (1045):
(10-5 1)
using
b7
1,2,3
(10-52)
(1053)
1,
2, 3
=
and have the form
pnj
(1054)
k
*_
SEC. 104.
ANALYSIS OF STRESS
247
These
equations apply for arbitrary strain and finite rotation. For small
strain, we neglect the change in dimensions and shape of the volume element. This assumption is introduced by taking
(1056)
Since the deformed unit vectors are orthogonal (toe 1), the Kirchhoff stresses now comprise a second-order cartesian tensor and they transform according
p2
x2
1,, =
4
oj dx2
dx2
dx1
p,,ds pn =
ij
c4 dx1
/
(1
e2)dx2
dx2
dx
(1 + 1)dx1
x1
to (1041).
For infinitesimal
and
In what follows, we will work with the Kirchhoff stress components to keep the treatment general. However, we will assume small strain.
t See Prob. 1016.
248
CHAP. 10
105.
A body is said to be elastic if it returns to its initial dimensions and shape when the applied forces are removed. The work done during the deformation process is independent of the order in which the body is deformed. We treat first an arbitrary elastic material and then specialize the results for a linearly elastic material. Our starting point is the first law of thermodynamics:
5W =
OVT
+ OQ
where OW = first-order work done by the forces acting on the body 0 VT = first-order change in the total strain energy (also called internal
energy)
V=
Y12'
.)
(1057)
The material is said to be hyperelastic (Green-type) when V is a continuous function. This requires
=
oek,
cejj ciek(
(1058)
By definition,
=
where
cetj
(1059)
is the first-order changet in due to an incremental displacement, M. Also, one can show that the first order work done by the force vectors
acting on the element is
OW =
.
+
dx2 dx3
(1060)
Equating OVT and OW leads to the general form of the stress-strain relation for a Green-type material,
=
(3
(1061)
tSeeProb. 1011.
See Prob. 1018. The forces are in equilibrium. is. they satisfy (1050).
SEC. 105.
249
This definition applies for arbitrary strain. Once V is specified, we can obtain
expressions for the stresses in terms of the strains by differentiating V. Since V is continuous, the stress-strain relations must satisfy (1058), which requires
(1062)
43e1j
In what follows, we restrict the discussion to small strain and a linearly elastic material, i.e., where the stress-strain relations are linear. We also shift from indicial notation to matrix notation, which is more convenient for this phase.
We list the stress and strain components in column matrices and drop the superscript k on the Kirchhoff stress components:
=
= {e11, e22, e33, 2e12, 2e23, 2e31}
(1063)
=
With this notation,
(1064)
(1065)
Since V is invariant under a transformation of reference frames, the transformation matrices are related by
1
(1066)
(1067)
where a contains the initial strains not associated with stress, e.g., strain due to a temperature increment, and A is called the material compliance matrix. We write the inverted relations as
= D(a
a)
(1068)
where D = is the material rigidity matrix. Equation (1062) requires D (and A) to be symmetrical. The elements of A are determined from material tests, and D is generated by inverting A. Substituting for in (1064), we obtain the form of the strain energy density for the linear case,
A'
V=
a)TD(a
a)
(1069)
Since V > 0 for arbitrary (E a), D and A are positive definite matrices.
There are 21 material constants for a linearly elastic Green-type material. The number of independent constants is reduced if the material structure
t See Prob. 106, 1013.
250
CHAP. 10
In what follows, we describe the transition from an anisotropic material to an isotropic material. A material whose structure has three orthogonal axes of symmetry is called orthotropic. The structure of an orthotropic material appears identical after a 1800 rotation about a symmetry axis. To determine the number of independent constants for this case, we suppose X1, X2, X3 are axes of symmetry and consider a 180 rotation about X2. We use a prime superscript to indicate the rotated axes. From Fig. 1013,
exhibits
= x1
= -x3
= x2
The stress and deformation quantities are related by (we replace 1 by I and 3 by 3 in the shear terms)
= a12 =
Y12 =
a12
Y12
= 1,2.3
a23
a23
a13 =
Y13
a13
Y13
Y23
Now, the stress-strain relations must be identical in form. We expand e = Acv', and substitute for using (b). Equating the expressions for a'
Fig. 1013. Rotation of axes for symmetry with respect to the X2-X3 plane.
tA
SEC. 105.
251
and
314(T12
ti15a33
a25a23
a24a12 a34u12
For (c) to be satisfied, the coefficients must vanish identically. This requires
434 = a15
a24 =
435
a34a350
The symmetry conditions require We consider next the expansions for a46 = a56 = 0. By rotating 1800 about X1, we find = a36 = a45 0 a16 = A rotation about the X3 axis will not result in any additional conditions. Finally, when the strains are referred to the structural symmetry axes, the stress-strain relations for an orthotropic material reduce to
a11
a12
444
-
a12
a22
0j3
a23
0
-
a1,
0
0
0 0
a12 a23
(733
(10-70)
0
Y31
a66
We see that A is quasi-diagonal and involves 9 independent constants. There is no interaction between extension and shear. Also, the shearing effect is
uncoupled, i.e., cr12 leads only to
An alternate form of the orthotropic stress-strain relations is
!(733
V32
a1 =
AT +
1
E2
-----a33
(1071)
a33
=
1
/13
AT +
i-
Y12 =
Y23
Y31
are shear moduli, Vjk are coupling coefficients, and AT is the temperature increment. The coupling terms are related by
E2
E1
E3
E1
E3
E2
(10-72)
252
CHAP. 10
If the stress-strain relations are invariant for arbitrary directions in a plane, the material is said to be transversely orthotropic or isotropic with respect to the plane. We consider the case where the X1 direction is the preferred direction, i.e., where the material is isotropic with respect to the X2-X3 plane. By definition, A is invariant when we transform from X1-X2-X3 to
This
c, ''12'-'31'--'
v32
v23
2(1 + v)
F
--
and
AT +
(a22
+
(1074)
= PAT +
= pAT +
1
1
va27)
y23
Yi2
2(l+v)
Y31
Lastly, the material is called isotropic when the stress-strain relations are Invariant for arbitrary directions, For this case, A = A' for arbitrary The relations are obtained by specializing (1074):
= p AT +
(at,
+ akk))
(1075)
2(1 + v) F
Note that now there are only two independent constants (F, v). The coupling coefficient, v,is called Poisson's ratio.
The inverted form of (1075) is written as
a=
a0
a0
+
(1076)
= =
+ 2G)pAT
Prob.
1022.
SEC. 106.
253
where
G=
A
shear
modulus =
yE
E
2(1
+ v)
(1077)
(1
+ v)(1 2v)
Since D must be positive definite, v is restricted to 1 < v < 1/2. The limiting case where v = + 1/2 is discussed in Problem 1024.
PRINCIPLE OF VIRTUAL DISPLACEMENTS; PRINCIPLE OF. STATIONARY POTENTIAL ENERGY; CLASSICAL STABILITY CRITERIA
106.
Chapter 7 dealt with variational principles for an ideal truss. For completeness, we derive here the 3-dimensional form of the principle of virtual
displacements, principle of stationary potential energy, and the classical stability criterion. The principle of virtual forces and stationary complementary energy are treated in the next section.
The principle of virtual displacements states that the Iirst-order work done is equal to the first oidcr work done by the internal forces acting on the restraints for an arbitrary virtual displacement of the body from an equilibrium position. f In the continuous case, the external loads and the internal forces are loading consists of body (b) and surface represented by the stress vectors. We follow the Lagrange approach, i.e., we work with Lagrange finite strain components (eJ,j, Kirchhoff stresses and external force measures per unit initial volume or area (b*, p*). This is consistent with our derivation of the equilibrium equations. Let Au denote the virtual displacement. The firstorder external work is
by the external forces
= =
1078
where fI is the initial surface area. The total internal deformation work is
obtained by summing the first-order work done by the stress vectors acting on a differential volume element. *
= =
dx2dx3
dx1 dx2 dx3
(1079)
Equating (a) and (b), we obtain the 3-dimensional form of the principle of
See Sec. 72.
See (1060).
254
CHAP. 10
virtual displacements,
5WD
=
dx1 dx2 dx3 +
(1080)
Requiring (1080) to be satisfied for arbitrary (continuous) is equivalent to enforcing the equilibrium equations. To show this, we work with the vector form and utilize the following integration by parts formula: t
=
where
dx2dx3
(10-81)
to the
is the direction cosine for the initial outward normal (n) with respect direction. Operating on the left-hand term and equating coefficients
in the volume and surface integrals leads directly to (1050) and (1054).
is independent of time, it can be interpreted as a variational principle for the displacements. The essential steps required for the truss formulation are described in Sec. 74. Their extension to a continuous body is straightforward. When the behavior is elastic,
=
Letting
VT
denote the total strain energy, the left-hand side of (1080) reduces to
+
where displacements are prescribed on
U1
on cd
(1082)
on
on
(1083)
We also consider the surface and body forces to be independent of the displacements. With these definitions, the principle of virtual displacements is transt See Prob. 1025.
SEC. 106.
255
formed to
= fl,, =
where
0
VT
(1084)
is the total potential energy functional. According to (1084), the displacements defining an equilibrium position correspond to a stationary value of the total potential energy functional. Note that this result applies for arbitrary strain and finite rotations. The only restrictions are elastic behavior
and static loading.
PH
Example 102
Direct methods of variational calculus such as Rayleigh-Ritz, Galerkin, weighted residuals, and others are applied to fl,, to determine approximate solutions for the displacements. In the Rayleigh-Ritz method, one expresses the displacements in terms of unknown parameters, q, and prescribed functions, x2, x3),
U1
where
forj =
>
1.
The displacement boundary conditions on fd are called "essential" boundary conditions. to a function of the q's. When the material is linearly Substituting for transforms elastic, V is a quadratic function of the strains. Then, V will involve up to fourth-degree reduces terms for the geometrically nonlinear case. If the behavior is completely linear,
H, =
q =
Const.
+ qTQ +
. .
.
(3N x 1)
K is symmetrical
256
CHAP. 10
Finally, requiring
Kq = Q
The strains are evaluated by operating on (a) and the stresses are determined from the
stress-strain relations.
Polynomials and trigonometric functions are generally used to construct the spatial distribution functions. The mathematical basis for direct methods is treated in numerous
texts (see Refs. 9, 10).
o2WE
>
for arbitrary Ad
is the second-order work done by the external forces where = during the incremental displacement, Ad, and WD = ( WD) is the second-
order work done by the internal forces acting on the restraints during the incremental deformation resulting from Ad. The form of the work terms for a continuous body are obtained by operating on (1078) and
= Sfl = =
=
Ad dx1 dx2 dx3 + j(
Ad
(1085)
Au1 dx1 dx2 dx3 + J$ Au, dx2 dx3 ie11 + dx2 dx3
If = 2WE for a particular Ad, the equilibrium position is neutral. The position is unstable if 2 WD < o2 Note that b, are itull vectors when the forces are prescribed.
= >
0
for arbitrary Ad
(1086)
value of the total potential energy. Bifurcation (neutral equilibrium) occurs when = 0 for some Ad, say Ada. If the loading is prescribed, and 2VT = 0 at bifurcation. The governing equations for bifurcation can be obtained by expanding This involves transforming the integrand of 2WD = by applying (1081). Since bifurcation corresponds to the existence of an alternate equilibrium position, it is more convenient to form the incremental equations directly. The equations for the case of linearly elastic material and prescribed external forces are listed below.
f See Sec. 76 for a derivation of the classical stability criterion.
See Probs. 1011, 1018.
SEC. 107.
257
+
2.
= 1,2,3
+
3.
Au1
J = 1, 2, 3
(1087)
Stress-Strain Relations
= D
4.
Strain-Displacement Relations
3 + AUJ, +
Au1
107.
Let u1 be the actual displacements in a body due to some loading and the geometrically linear strain measures corresponding to u1. The strain and displacement measures are related by
u1=fl
Once the strains are known, we can find the displacements by solving (a) and enforcing (b). The principle of virtual forces is basically a procedure for determining displacements without having to operate on (a). It applies only for linear geometry. We developed its form for an ideal truss in Sec. 73. We will follow the same approach here to establish the three-dimensional form. The essential step involves selecting a statically permissible force system, i.e., a force system which satisfies the linear equilibrium equations. For the continuous case, the force system consists of stresses, surface forces, on Static permissibility requires and reactions, on
Ac31,3 = 0
= =
on on
(1088)
If we multiply e13 by Ac13, integrate over the volume using (1081), and note the
static relations, we obtain the following identity,t
Acr13 dx1 dx2 dx3
f See Prob. 1026.
u1
+ $ Th
(1089)
0,,
258
CHAP. 10
which is referred to as the principle of virtual forces (or stresses). This result is
in the direction defined by is Suppose the translation at a point Q on desired (see Fig. 1015). Let d0 be the displacement. We apply a unit force at Q in the tq direction and generate a statically permissible stress field.
(1) 1q at
point Q
Acr
and
The integral on
(1090)
A second application is in the force method, where one reduces the governing
involving only force unknowns. We start by expressing the stress field in terms
of a prescribed distribution
is a particular solution of the equilibrium equations which satisfies the boundary conditions on
where
=
and
satisfies
Thu
on
(10-92)
= =
0 0
on
on
(1093)
Stress fields satisfying (1093) are called seljequilibrating stress fields. For the
ideal truss, a- corresponds to the forces in the primary structure due to the prescribed loading and ? represents the contribution of the force redundants.
SEC. 107.
259
The governing equations for the force redundants were obtained by enforcing
geometric compatibility, i.e., the bar elongations are constrained by the requirement that the deformed bar lengths fit in the assembled structure.
in Prob. 1010. One can also obtain these equations with the principle of
virtual forces by taking a self-equilibrating force system. Letting Aox, Apc denote the virtual stress system, (10--89) reduces to
(1094)
The compatibility equations are determined by expressing in terms of stress functions and integrating the left-hand term by parts. We illustrate its application to the plane stress problem.
Example 103 If the stress components associated with the normal direction to a plane are zero, the
stress state is called planar. We consider the case where
+
012 5 +
= 033 = 0. The
a,,2o21
z,,2a22
The stress field, oi,, must satisfy (a) with h1 = h2 = 0 and also p,,1 = = 0 on We can satisfy the equilibrium equations by expressing in terms of a function, follows:t 033 = = =
The boundary forces corresponding to
Pa
as
are
OS
pC
if
CS
Ps
0
,i
+ a2,
712,
260
CHAP. 10
+
which is actually a continuity requirement
U1 122
if
ApC
We express the stress matrix in terms of prescribed stress states and unknown
parameters, a1,
= =
+
+
(1
+ (12(l)2
'
+ 04,,
where
states, i.e., They satisfy the homogenous equilibrium equations and boundary conditions on The corresponding surface forces arc
p
= p + 0101 + 0209 + = p (i = 1, 2, = 0
(i =
1
1, 2,
., r) results in r
(1097)
1, 2,...,r
In order to proceed, we need to introduce the material properties. When the material is linearly elastic,
+ Ai'=
and the equations expand to
d1
i,j =
r$J1T(a
1,
2,....r
(1098)
f,j =
d, =
One should note that (1097) are weighted compatibility conditions. The true stresses must satisfy both equilibrium and compatibility throughout the
t See Prob. 1027.
SEC. 107.
261
the corrective stress field since it is required to correct the compatibility error due to For completeness, we describe here how one establishes a variational principle for Our starting point is (1094) restricted to elastic behavior. We define = according to = c5V* = (1099)
domain. We call
and call V* the complementary energy density. The form of V* for a linearly elastic material is = (10100) +
By definition, V* complements V, i.e.,
V+
Then, letting
= =
cjx1 dx2 dx3
(10101) (10102)
0
*
for arbitrary = $$
(10103)
This form is called the principle of stationary complementary energy and shows
that the true stresses correspond to a stationary value of Since is linear in the second variation of reduces to
(52fl
= =
A,
(10104)
represents the tangent compliance matrix. Now, must be positive definite in order for the material to be stable.t Then, (5211. > 0 for arbitrary and we see that the solution actually corresponds to a relative minimum
where value of The approximate method described earlier can be applied to 11g. Substituting for given by (1095) converts to a function of the stress parameters When the material is linearly elastic, (a1. a2, . , ar).
.
.
H, =
ard + const
(10106)
The equations for the stress parameters follow by requiring H. to be stationary for arbitrary (511, = ci) = 0
fa=d
A,
(10107)
The classical stability criterion specialized for elastic material and linear geometry requires SCTD, & > 0 for arbitrary Sc which, in turn, requires D, to be positive definite. Since D1', it follows that A must be positive definite for a stable material.
262
CHAP. 10
Operating on
c52fl = LtaTfLui
(10108)
CRANDALL, S. J., and N. C. DAHL: An Introduction to the Mechanics of Solids, McGraw-Hill, New York, 1959.
of Deformable Solids. Addison-Wesley, Reading, Mass., 1965. WANG, C. T.: Applied Elasticity, McGraw-I-jill, New York, 1953. TIMOSHENKO, S. J., and J. N. GooDiag: Theory of Elasticity, 3d ed., McGraw-Hill, New York, 1970. SOKOLNIKOFF, I. S.: Mathematical Theory of Elasticity, 2d ed. McGraw-Hill. New York, 1956. FUNG, Y. C.: Foundations of Solid Mechanics, Prentice-Hall, 1965. LEKIINITSKU, S. G. Theory of Elasticity of an Anisotropic Elastic Body, Holden-Day, San Francisco, 1963. WAsmzu, K. Variational Methods in Elasticity and Plasticity, Pergamon Press, 1968. HLDEBRAND, F. B.: Methods of Applied Mathematics, Prentice-Hall, 1965. CRANDALL, S. J.: Engineering Analysis, McGraw-Hill. New York, 1956.
BISPLINGHOFF, R. L., MAR., J. W., and T. H. H. PlAN:
5. 6. 7. 8. 9.
10.
PROBLEMS
101.
where
+ Urn. k)
Let f be a continuous function of x1, x2, x3. Establish the transJbk where
formation laws for and (3Xk. 103. Establish the transformation law for tensors. 104. Prove that
eJk =
are cartesian
Jk)
(If)
P.
P.
105. Equations (1019) are the strain transformation laws. Since is a symmetrical second-order cartesian tensor, there exists a particular set of directions, say Xi', for which is a diagonal array. What are the strain components for the frame? Consider a rectangular parallelepiped having sides dXy in the undeformed state. What is its deformed shape and relative change with respect to its initial volume? Specialize the expression for in volume, for small strain. Then determine for the initial (Xi) directions and small strain. Finally, show that r.., is invariant.
PROBLEMS 106.
263
(a)
(b)
Specialize (1019) for small strain and write out the expressions for
in terms of ei, 62, .
.
P13
P12, P23, y31}. We can express the strain trans= formation (small strain) as =
Let
x2
107.
for the
2Efk
thik
Determine the expression for EJk in terms of the displacements. Compare the
result with (1021).
108.
plane (83 = P13 = P23 = 0). Let 6b, tions defined below and let 6N = {8a, 6b,
Consider the case of two-dimensional deformation in the X1-X2 be the extensions in the a, b, c direcWe can write
= BE C=
(a) (b)
(c)
= 90c. = 120. (d) Extend (a) to the three-dimensional case. Consider six directions having direction cosines GJ2, with respect to X1, X2, X3. Can we select the six directions arbitrarily?
264
CHAP. 10
Prob. 108
xz
-a
109.
=
Rather than work with
C3
= eti + &22 + +
=
=
where
is the deviator strain tensor. and (a) Write out the expanded form for (b) Determine the first invariant of of ejj.
1010.
(a)
+
where
=
CX,, cXj, 8X,,, (CII,,
(?X,,
eflk =
ek,
2 \CXk
and k, in, n range from I to 3. This expression leads to six independent conditions, called geometric compatibility relations, on the strain
measures.
(b)
Show that for two-dimensional deformation in the X1-X2 plane = 0; this called plane strain) there is only one com= 813 =
patibility equation, and it has the following form:
22 + 83 11 =
Is
Y12. 12
= constant
PROBLEMS
1011.
265
To analyze geometrically nonlinear behavior, one can employ an incremental formulation. Let represent the displacement increment and Ae1k the incremental strain. We write + =
where
contains linear terms (Aug) and eJk involves quadratic terms. The 5-symbol denotes the first-order change in a functional and is called the variational operator (see Ref. 8). We refer to 5e as the first variation of e. Determine the expressions for
1012.
Let
= Determine the general expression for Then specialize it for small strain. 1013. The several parts of this question concerns stress transformation. in terms of (a) Starting with (i0--41). write out the expressions for all, a22, . , = stress matrix. We express the (b) Let a22, a33, ai2, = stress transformation as a matrix product.
.
= (1 +
a' = T,a
Develop the form of T,, using the results of part a.
(c)
Evaluate 1',, in terms of cos (9, sin (9 for the axes shown.
x2
Prob. 1O13
x,t
xl
(d)
Plane stress refers to the case where a13 = with reduced stress and strain matrices,
Er
a23
a33
0.
We work
{a11, a22,
=
and write the transformations in the same form as the three-dimensional case:
a' =
a'
266
CHAP. 10
1014. This question develops a procedure for generating self-equilibrating stress fields. (a) Expand the linear equilibrium equations, (1049) and (1050). (b) Specialize the equilibrium equations for plane stress (a13 = a23 =
(c)
= 0). Suppose we express the two-dimensional stress components in terms of a function = as follows:
a33
b1
dx1
fx7 b2
dx2
= 1//,12
The notation for body and surface forces is defined in the following
sketch.
Prob.
x2
1014
x1
Verify that this definition satisfies the equilibrium equations in the interior. Show that the expressions for and P2 ifl terms of derivatives
b1
dx1
b2
T t/"1
dx2
1015.
a22
+ a33)
we
PROBLEMS
267
where
=
and
(a) (b)
ijOrn
is the deviator stress tensor. Write out the expanded forms for and Determine the first invariant of 10-46. Establish the stress-equilibrium equations for small-finite rotation and small strain.
1017.
Starting with (1052), (1055) specialized for small strain, establish can
Au, Ab*, and the incremental equilibrium equations in terms of Group according to linear and quadratic terms. Specialize these equations
for the case where the initial position is geometrically linear, i.e., where approximate with in the incremental equations. 1018. Prove (1060). Hint:
= /=
1019.
+ P.k
where
D11 = E1/C3
D12 D13
C4D11
and
C1 = C2 =
v21
+ v31v32(E2/E3)
E2C1
C4 =
v31
"32
and
= 0) = = Consider 2 sets of orthogonal directions defined by the unit vectors The stress-strain relations for the two frames are
+ (a)' + A'&
Express A' in terms of A and Also determine D'. 1O--21. Consider the three-dimensional stress-strain relations defined by
(1071).
(a)
= 0).
268
CHAP. 10
(b)
Let
a22, cri2}
C= C=
62, Y12}
0
1
G
(1
E2
x2
xI
1022. Verify (1073). Start by requiring equal properties for the X2 and X3 directions. Then introduce a rotation about the X1 axis and consider the Isotropy in the X2-X3 plane requires expression for
Y23
=7
I.
023
1023. Verify that the directions of principal stress and strain coincide for an isotropic material. Is this also true for an orthotropic material?
1024.
2Ge11
where
(a)
and 1015
Show that
Ka,,
+ a0
PROBLEMS
269
Show that
=
(c)
= =
(d)
for the isotropic case. Determine and When v = We must work with 7 stress measures ('u' Urn) = and the mean stress has to be determined from an equilibrium consideration. Summarize the governing equations for the incompressible
case.
Prove (l0--81) for the two-dimensional case. Is this formula restricted to a specific direction of integration on the boundary? Does it apply for a multi-connected region, such as shown in the figure below?
.1025.
Prob.
1025
1026.
Refer to Example 103. Express (g) in terms of material to be orthotropic. 1028. Verify that the stationary requirement
1027.
=0
where
for arbitrary
dx2
1u,,,,
= complementary energy density (initial volume) = prescribed force measures (initial dimensions) leads to the complete set of, governing equations for an elastic solid, i.e., stress equilibrium equations 1. 2. stress-displacement relations 3. stress boundary conditions on 4. displacement boundary conditions on 5. expressions for the reaction surface forces on
270
CHAP. 10
Transform HR to by requiring the stresses to satisfy the stress displacement relations. Hint: Note (10101). (b) Transform 11R to by restricting the geometry to be linear = and (ui, + and requiring the stresses to satisfy the stress equilibrium equations and stress boundary conditions on Hint: Integrate by parts, using (108 1).
(a)
1029.
Interpret (1090) as
dQ ==
11
St.
Venant Theory of
of
Prismatic Members
111.
A body whose cross-sectional dimensions are small in comparison with its axial dimension is called a member. If the centroidal axis is straight and the the member shape and orientation of the normal cross section are is said to be prismatic. We define the member geometry with respect to a global reference frame (X1, X2, X3), as shown in Fig. 111. The X1 axis is taken to coincide with the centroidal axis and X2, X3 are taken as the principal inertia directions. We employ the following notation for the cross-sectional properties: A = if dx2 dx3 = dA
12 Sj(x3)2 dA
13 = fl(x2)2 dA
Since X2, X3 pass through the centroid and are principal inertia directions, the centroidal coordinates and product of inertia vanish:
'23
jJx2x3 dA = 0
One can work with an arbitrary orientation of the reference axes, but this will
complicate the derivation. St. Venant's theory of torsion-flexure is restricted to linear behavior. It is an
272
CHAP. it
Chapter 13, we modify the St. Venant theory to account for displacement restraint at the ends and for geometric nonlinearity.
x2
F3
The distribution of surface forces on a cross section is specified in terms of its statically equivalent force system at the centroid. Figure 111 shows the Stress components on a positive face. We define M.. as the force and moment vectors acting at the centroid which are statically equivalent to the distribution of stresses over the section. The components of F.., M are called stress resultants and stress couples, respectively, and their definition equations are
F1 = ffcrij c/A M1 =
M2 = M3 =
F2
JJx3crj1 dA
c/A
F3 =
JJcc13 c/A
JJ(x2cr13 x3c12)dA
(113)
dA
The internal force and moment vectors acting on the negative face are denoted
byF_,M_. Since
F_ =
M_ =
(114)
it follows that the positive sense of the stress resultants and couples for the
negative face is opposite to that shown in Fig. 11 1. We discuss next the pure-torsion case, i.e., where the end forces are statically equivalent to only M1. We then extend the formulation to account for flexure
SEC. 112.
273
and treat torsional-flexural coupling. Finally, we describe an approximate procedure for determining the flexural shear stress distribution in thin-walled
sections.
112.
Consider the prismatic member shown in Fig. 112. There are no boundary forces acting on the cylindrical surface. The boundary forces acting on the end
cross sections arc statically equivalent to just a twisting moment M1. Also,
there is no restraint with respect to axial (out-of-plane) displacement at the ends.
The analysis of this member presents the pure-torsion problem. In what follows, we establish the governing equations for pure torsion, using the approach originally suggested by St. Venant.
Rather than attempt to solve the three-dimensional problem directly, we impose the following conditions on the behavior and then determine what problem these conditions correspond to.
1.
2.
Each cross section is rigid with respect to deformation in its plane, i.e., = 723 = 0. = Each cross section experiences a rotation w5 about the X3 axist and
an out-of-plane displacement u1.
These conditions lead to the following expansions for the in-place displacements:
112 = C01X3
03
+W3.3.2
(115)
Y23
0
(116)
712
713
U12 +
=
U5,
012
+ U3,
05,3 + X20)1, 1
t Problem 11i treats the general case where the cross section rotates about an arbitrary point.
274
CHAP. 11
Now, the strains must be independent of x1 since each cross section subjected to the same moment. This requires = const = k1 = u1(x2, x3)
is
We consider the left end to be fixed with respect to rotation and express
co1,u1aS
=
U.1 =
k1x1
(118)
where
x3) defines the out-of-plane displacement (warping) of a cross = section. The strains and stresses corresponding to this postulated displacement behavior are = 0 = C3 Cl =
712
=
=
x3)
713
+ x2)
a33
and
a11 =
a22
O'23 = 0
a12 =
U13 =
Gy12
Go'13
2 x3)
a12(x2, x3)
(1110)
+ x2)
a1 3(x2, x3)
We are assuming that the material is and there are no initial strains. One step remains, namely, to satisfy the stress-equilibrium equations and stress boundary conditions on the cylindrical surface. The complete system of linear stress-equilibrium equations, (1049), reduces to
U21,2 +
(1111)
Substituting for the shearing stresses and noting that Gk1 is constant lead to the differential equation
(11-12)
which must be satisfied at all points in the cross section. The exterior normal n for the cylindrical surface is perpendicular to the X1 direction. Then 0, and the stress boundary conditions, (1049), reduce to
Pfli =
+
is
=0
+ x2) =
(on S)
(1113)
(1114)
SEC. 112.
275
b, is known, we determine the distribution of transverse shearing stresses from (1110). Note that depends oniy on the shape of the cross section.
F2 =
This requires
dA =
F3 = J$a13 dA
J J OX2
JJ ('X3
if
Green's theorem,
JJV2VJdA
dA =
(IS
dA leads to
0X2
+
1116
.1
If
0),
dS =
#(XH2x3
is specified on the boundary, we cannot apply (1115) directly Since to (b). In this case, we use the fact that = 0 and write cxi
Integrating (e),
ax2
ax2j
ox3 \
ax31
(j=2,3)
(j=2,3)
and then substituting for the normal derivative, verifies (h). The constant k1 is determined from the remaining boundary condition,
J$(x2c13
x3c12)dA
(1117)
276
CHAP. 11
Gk1J
where J is a cross-sectional property,
if
X3
dA
Displacements
=
02
U3 = W1X2
= k1x1
k1
(if
2.
Stresses
M3
J
0j3 = H J
3.
A'11
\(;X2
(1120)
+ X2
Governing Equations
mA:
on S:
It is possible to obtain the exact solution for for simple cross sections. The procedure outlined above is basically a displacement method. One can also use a force approach for this problem. We start by expressing the shearing stresses in terms of a stress function so that the stress-equilibrium equation (Equation 11li) is identically satisfied. An appropriate definition is
012
013
X3
(1121)
0lv =
SEC. 11-2.
277
Taking S 900 counterclockwise from the exterior normal direction, and noting
0,
= const on S
We establish the differential equation for t/i by requiring the warping function be continuous. First, we equate the expressions for a in terms of t/i and
a12 =
a13 =
Now, for continuity,
M1
x3)
= =
+ x2)
=
It is convenient to express t/.'
as
(1124)
aret
M1 dt/
(11--25)
a12 = =
a 13
j
(mA)
(on boundary S1)
(1126)
and
=
tJi =
Substituting (1125) in the definition equation forM1 leads to the following expression for J:
cr7
JJ \.
that
CX3J
dS =
A1
(b)
C1
= const
t Equations (1126) can be interpreted as the governing equations for an initially stretched membrane subjected to normal pressure. This interpretation is called the "membrane
See Ref. 3.
The S direction is always taken such that n S has the same sense as X2 X3. Then, the + S direction for an interior boundary is opposite to the + S direction for an exterior boundary
since the direction for n is reversed. This is the reason for the negative sign on the boundary integral.
278
CHAP. 11
we can write
where
J=
dA +
(1127)
= 0 on the exterior boundary. To determine the constants C, for the multiply connected case, we use the is continuous. This requires fact that
Js ('IS
dS
0
(1128)
x2
SEC. 112.
279
is
given by
Yis =
Using (119), we can write (a) as
ct52y12
+ 0t53y13
Yis =
2+
k1
3 xacls2
+
(11-29)
where p is the projection of the radius vector on the outward normal.t The magnitude of p is equal to the perpendicular distance from the origin to the tangent. Integrating between points P, Q, we obtain
=
where
APQ
+ 2APQ)
(1130)
r50
J
dS =
Finally, taking P =
5dS = 2k1A5
(1131)
we can (11-32)
Note that the +S direction for (1132) is from .X2 toward X3. Also, this result is independent of the location of the origin. Instead of using (119), we could have started with the fact that the cross section rotates about the centroid. The displacement in the + S direction follows from Fig. 114:1
u,5 =
x
is)
Us
=
+
w1p
k1x1p
(1133)
Substituting for
in
Yss =
(1134)
M5
(1135)
t This interpretation of p is valid only when S is directed from X2 to X3, i.e., counterclockwise
for this case. See Prob. 1114 for an alternate derivation.
This development applies for arbitrary choice of the +S direction. The sign of p is positive if a rotation about X1 produces a translation in the +S direction. Equation (1129) is used to
determine the warping distribution once the shearing stress distribution is known. See Prob. 114.
280
CHAP. 11
in (1132), we obtain
=
.3 s
(1136)
where n is the outward normal, A5 is the area enclosed by S, and the + S sense is from X2 to X3. This result is valid for an arbitrary closed curve in the cross section. We employ (1136) to determine the values of 17 at the interior boundaries of a multiply connected cross section. It is of interest to determine the energy functions associated with pure torsion. When the material is linearly elastic and there are no initial strains, the
V dA
(1137)
V=
Substituting for Y12' '/13,
V=
X3)
+(
x2)j
2
V=
Since
(1138)
V,
WI
xl
dx1
Instead of integrating the strain-energy density, we could have determined the work done by the moments acting on a differential element. Consider the element shown in Fig. 115. The boundary forces acting on a face are statically equivalent to just a torsional moment. Also, the cross sections are rigid in
SEC. 11S.
281
their plane and rotate about X1. The relative rotation of the faces is
dx1
5WE =
Now,
5jJ
dx2dx3 = Vdx1
= M1 =
GJk1
V=
113.
APPROXIMATE SOLUTION OF THE TORSION PROBLEM FOR THIN-WALLED OPEN CROSS SECTIONS
We consider first the rectangular cross section shown in Fig. 116. The exact solution for this problem is contained in numerous texts (e.g., see Art. 53 of Ref. 1) and thcrefore we will only summarize the results obtained.
x3
dl 2
d, 2
282
CHAP. 11
0
When t d, the maximum shearing stress (points 5, 6). The exact expressions are
occurs at x2 = t/2, x3 =
J = K1
(11-41) =
where K2t
dt3
K1 = K2 =
192 (t'\
1
I
=
tanh
8
1
(2n+1)2 cosh
A,,
2n+1 Id
Values of K1, K2 for d/t ranging from 1 to 10 are tabulated below:
d/t
1
K1
K2
0.422
.687 .789
.843
2
3
4
5
.873
10
0.936
If t d, we say the cross section is thin. The approximate solution for a thin rectangle is J 4dt3
(113
2-x2
2Gk1x2
(11-42)
x2x3
(t)2
(We
take d/t = in the exact solution.) The shearing stress across the thickness and
M1
varies linearly
3M1
Since the stress function approach is quite convenient for the analysis of thin-walled cross sections, we illustrate its application to a thin rectangular
SEC. 113.
283
section. Later, we shall extend the results obtained for this case to an arbitrary thin walled open cross section. The governing equations for a simply connected cross section are summarized below for convenience (see (1126),
cross
(1127)):
(in A)
=
J =
(1A
where the S direction is 900 counterclockwise from the is direction.t Since t is small and a12, the shearing Stress component in the thickness direction, must
_:k,1,
Fig. 117. Warping function for a rectangular cross section. vanish on the boundary faces, it is reasonable to assume
dx2
M1 = ----- = 2x2
t This applies for X3 counterclockwise from X2. The general requirement is the n S sense must coincide with the X2-X3 sense.
284
CHAP. 11
This will lead to a12 0 near the ends, but will have a negligible effect on J Actually, the moment due to the approximate linear expansion for and is equal to only one half the applied moment:
I't/2
d
J
x2a13 dx, =
P4
/1
dt3) =
The corrective stress system (a12) carries M1/2. This is reasonable since, even is small in comparison to amax, its moment arm is large. though
-f-s
t(s)
J=
t3 dS
(1143)
M1
a15, ma, =
= Gkitrnax
SEC. 113.
285
The
results for a single thin rectangle are also applied to a cross section
t1 denote the length and thickness
(1144)
J=
ing (1144), we obtain
3 1'ff + w4v
..i
Asan illustration, consider the symmetrical section shown in Fig. 119. Apply3
The
In general, there is a stress concentration at a reentrant corner (e.g., point A in Fig. 119) which depends on the ratio of fillet radius to thickness. For the case
bi
I
Fig. 119. Symmetrical wide-flange section.
(Iw
4rf/
(1146)
is the fillet radius and 0rn is given by (1445), gives good results for where rf/t < 0.3. The stress increase can be significant for small values of rf/t. For example, for Tf = 0,lt. Numerical procedures such as finite differences or the finite element must be resorted to in order to obtain exact solutions for irregular sections.
-
286
CHAP. 11
114.
APPROXIMATE SOLUTION OF THE TORSION PROBLEM FOR THIN-WALLED CLOSED CROSS SECTIONS
The stress function method is generally used to analyze thin-walled closed cross sections. For convenience, the governing equations are summarized below
(see (1126), (1127), (1136)):
2
(in A)
ci
J
dA +
(on the exterior boundary) (on the interior boundary, S,) = area enclosed by
S
on
dS
We consider first the single cell shown in Fig. 1110. The curve defines the centerline. Since there is an interior boundary, we have to add a term
n
S.
Sect. E-E
2n\
(1147)
where
SEC. 114.
287
satisfies
C1
at
(a)
and substituting (b) in the expressions for the shearing stress components lead to
M1 /
+ C1\ 7)
(1148)
cr?5 + The tangential shearing stress varies linearly over the thickness and its average We let q be the shear stress resultant per unit length along S, value is positive when pointing in the + S direction,
1/2
cr15
(1149)
we find
(1150)
J 1/2
The additional shearing stress due to the interior boundary (i.e., closed cell) corresponds to a constant shear flow around the cell. One can readily verifyt that the distribution, q = const, is statically equivalent to only a torsional
moment,
given by
=
The torsional constant is determined from
(1151)
J
Substituting for
dA + 2C1A1
M1/Gk1
(a)
j = Jo +
4
t3 dS
(1152)
Equation (a) was established by substituting for the shearing stresses in terms in the definition equation forM1 and then transforming the integrand. We could have arrived at (1152) by first expressing the total torsional moment as
M1 =
See Prob. 11-5.
(1153)
288
CHAP. 11
we write
M1 = GkIJ
Then,
it'll =
Gk1J
= Gk1J'
(1154)
(1155)
J=
and it follows that
Jo
J0
+ JC
Jc
(11-56)
as
= Mu/(M1/J) =
q/(M1/J)
(1157)
This result shows that we should work with a modified shear flow,
C
(1158)
rather than with the actual shear flow. Note that C C1 for the single cell. It remains to determine C1 by enforcing continuity of the warping function
on the centerline curve. Applying (1132) to
we have
(1159)
=
Substituting for
q/t
leads
M1C1
(1160)
One should note that C1 is a property of the cross section. Once C1 is known, we can evaluate .J from (1152) and the shearing stress from
M1 (
+ -i,)
(1161)
Example 111
Consider the rectangular section shown. The thickness is constant and a, b are centerline properties are dimensions. The various
CdS
2(a+h)
Cl =
SEC. 11-4.
We express J as
289
=
For this section,
1 (r'Y / (t'Y
\\h
We consider a >
b.
Then,
Jo
J'
Jo vs.
= 01
c< b.
Fig. Elii
b H r+tb
-
+s, q
t\,
(I
h'\t
(t
i.e., we can take
We
consider next the section shown in Fig. 11li. Rather than work with
it is more convenient to work with the shear flows for the segments. We
number the closed cells consecutively and take the + S sense to coincide with the X2-X3 sense. The +S sense for the open segments is arbitrary. We define q3 as the shear flow for ellj and write
(1162)
Note that is the value of on the interior boundary of cell j and the shear flow is constant along a segment. The total shear flow distribution is obtained
290
CHAP. 11
x3
q2. S2
Fig. 1111. Cross section consisting of closed cells and open segments; A, and A,, are centerline areas.
by superimposing the individual cell flows. Then, the shear flow in the segment common to cells i andj is the difference between qj and q1. The sign depends on the sense of S.
q=
q2
=
forS2
C2)
for S1
(1163)
q=q2q1
The shearing stress is assumed to vary linearly ovcr the thickness. For convenience, we drop the subscripts on and write the limiting values as
cr = a + Cr"
where
cr=1-t
It remains to determine C1, C2, and J.
We have shown (see (1155)) that
M1 /Cnet
(1164)
J=
and
Jo + Jc
(a)
=
We determine J from
(h)
(1165)
segments
SEC. 114.
291
Substituting for
MI =
in (b) leads to
-
2qjA1 + 2q2A2
+ C2A2)
+ A2C2)
(1166)
The constants are obtained by enforcing continuity of on the centerline of each cell. This can also be interpreted as requiring each cell to have the same twist deformation, k14
I = 1,2
Substituting for q in terms of C and letting
(11-67)
dS
JS, t
a22
dS
Cads
.J52
Jc
where a12 involves the segment common to cells 1, 2, the continuity equations
(1169)
2A2
C2, then determine f with (1166), We solve this system of equations for and finally evaluate the stresses with (1164). We can represent the governing equations in compact form by introducing matrix notation. The form of the equations suggests that we define
[a11
a121
a22j
(1170)
= 2ATC
(117.!)
aC
2A
CTaC
The complementary energy per unit length along the centroidal axis is defined
11
2
292
CHAP. 11
Since
ais varies linearly over the thickness, the open and closed stress dis=
+
2G
toq
(11-72)
aq)
Examp'e 112
The open-section torsional constant for the section shown is
2(b
+d+
+ htfl
(a)
=
a11 =
012
hd
A2=hb
1(h + 2d) +
11
t2
=
= tl
a22
+ 2b) +
t2
+2
C2
dt1
c1 +
C2 = 2bt1
J=
Jo + Jo
M1 (C1 (k +
M1 /C1 C2
J
M1 (C2
t2
+t2
=7
=
M1
t3
+ t1
____ _____________
SEC. 115.
293
Fig. E112
t3
tl
03
Ii
I
1
I
032 A2
X3
a'
M1
When
d=
b.
Cl =
C2
s
2bt3
1 + 213
and the section functions as a single cell with respect to shear flow.
115.
Consider the prismatic member shown in Fig. 1112. There are no boundary
x2
_____
if
xI
P2
I +S
on the cross section at x1 = L is statically equivalent to a single force P212, acting at the centroid. Also, the end cross sections are not restrained against displacement. In what follows, we describe St. warping, i.e., Venant's torsion-flexure formulation for this problem. Later, in Chapter 13,
we shall modify the theory to include restraint against warping.
294
CHAP. il
We start by postulating expansions for the stresses. The stress resultants and couples required for equilibrium at x1 are
=A4,r0
F2 = P2 M3 = P2(L
x1)
Introducing (a) in the definition equations for the stress resultants and couples leads to the following conditions on the stresses:
=
,
S$(x2a13 x3cr12)dA
0 P2
The expansion,
M3
13
13
dA =
13
The last three conditions (i.c., F2, F3, M1) require a12, a13 to be independent of x1. This suggests that we consider the following postulated stress behavior:
cru =
a1 2
--x2 =
13
P2
x1)x2
(11 73)
13
a1 2(x2, x3)
a33
Introducing (1173) in the stress-equilibrium equations and stress boundary conditions for the cylindrical surface leads to
a21,2 + a31,3 +
2a21 +
P2
13
(mA)
(on S)
(1174)
=0
At this point, we can either introduce a stress function or express (1174) in terms of a warping function. We will describe the latter approach first. The displacements can be found by integrating the stress-displacement relations. We suppose the material is linearly elastic, isotropic with respect to the X2-X3 plane, and orthotropic with respect to the axial direction. This is a convenient way of keeping track of the coupling between axial and in-plane
SEC. 115.
295
u1
P2
E113
V1
E1
(L
xj)x2
x1)x2
= u2,2 =
V1P2
(L El3
v1P2 7(L
63 = u33 =
'Y12
v1
=
=
I
LI3
xj)x2
= u1, 2 + u2,
= function ofx2, x3
Y13
Y23
U1 3 + U3, 1 =
= U2,3 + U3,2
0
= function ofx2, x3
= =
E1 13
(Lx1
+ f1(x2, x3)
(b)
v1P2 v1P2
(L
+ .f2(xj, x3)
x1)x2x3
u3 = - (L
El3
x2)
The functions f1, f2, f3 are determined by substituting (b) in the last three equations. We omit the details and just list the resulting expressions, which
involve seven constants:
= f2 =
C1
+ C5x2 + C6x3 +
x3)
(c)
C2
C5x1
+ C4x3 k1x1x3
v1P2
Xi.
+ --7-(L
2E,iJ 3
C3 C5x1 C4x2
+ k1x1x2
The constants C1, C2, ..., C6 are associated with rigid body motion and k3 is associated with the twist deformation.t We consider the following displacement boundary conditions:
1.
u1=u2=u3==0
2.
at(0,0,0)
at(0,0,O)
=
fSee Eq. (115).
u3,1
296
3.
CHAP. Ii
A line element on the X2 axis at the origin is fixed with respect to rotation in the X2-X3 plane: at (0, 0, 0) u2,3 = 0
(Lxj
+ 4)(x2, x3)
=
Vj
2
(1175)
El3
x1)x2x3 + k1x1x2
One step remains, namely, to satisfy the equilibrium equation and boundary condition. The transverse shearing stresses are given by
1
41,2 k1x3
v1P2
(7j3
4),
+ k1x2
2L13 v1P2
x2) (1176)
LI3
x2x3
Substituting for the stresses in (1174), we obtain the following differential equation and boundary condition for 4):
P2f'2v1
l\
(mA)
v1P2 C
12) +
as
- 77
= kjq5t
(11-78)
where
is the warping function for pure torsion and 4)2. and 4)2d are harmonic functions which define the warping due to flexure. Substituting for 4) leads to the following boundary conditions for 4)2. and
2
2
2
)
=
0
(1179)
SEC. 115.
297
therefore the formulation is consistent. Terms involving vj/E are due to in-plane deformation, i.e., deformation in the plane of the cross section, and setting v1/E = 0 corresponds to assuming the cross section is rigid. Then, defines the flexural warping for a rigid cross section and represents the correction due to in-plane deformation. The shearing stress is obtained by substituting for in (1176). We write the result as (j = 2,3) (11SO) + 01j,r + 01j =
and
where crU, is the pure-torsion distribution and butions corresponding to and 42d:
r,
d are
flexural distri-
2
2
P2
x2x3)
ment,
G1k1J.
J$a12,a dA
dA 0 dA 0
(1182)
Note that the shear stress due to in-plane deformation does not contribute
to P2. The total torsional moment consists of a pure torsion term and two flexural
terms,
M1 = G1k1J +
S2r
S2d
+ X24)2r 3)dA
X34)2a,2)dA
+ X242d,3
Since
and depend only on the shape of the cross section, it follows that and S2d are properties of the cross section. For convenience, we let
X3
1
13\
=
(1184)
J
Now,
tSeeProb. IllO.
298
CHAP. 11
defines the location of the resultant of the flexural shear stress distribution with respect to the centroid. The twist deformation is determined from
k1
(1186)
where M1 is the applied torsional moment with respect to the centroid. If P2 is applied at the centroid, M1 = 0, and k1 =
The cross section will twist unless
= 0. Suppose P2 has an eccentricity e3. In this case (see Fig. 1113), M1 = e3P2, and
k1
P2
e3)
Now, to Whether twist occurs depends on the relative eccentricity, e3 find x3, one must determine S2. and S2d. This involves solving two secondorder partial differential equations. Exact solutions can be obtained for simple cross sections. In the section following, we present the exact solution for a rectangular cross section. If the section is irregular, one must resort to such numerical procedures as finite differences to solve the equations. In Sec. 117, we describe an approximate procedure for determining the flexural shear stress distribution in thin walled cross sections. Suppose the cross section is symmetrical with respect to the X2 axis. Then, is an odd function of x3. The form of the is an even function of x3 and boundary conditions (1179) requires and to be even functions of x3
SEC. 115.
299
this case. Finally, it Ibilows thatt S2,. = 0 and S2d = 0. Generalizing this result, we can state: The resultant of the shear stress distribution due to fiexure in the is an axis of symmetry direction passes through the centroid when for the cross section.
for
x3
Shear center
We consider next the case where the member is subjected to P2, P3 and at the right end (see Fig. 1114). The governing equations for the P3 loading can be obtained by transforming the equations for the P2 case according to
>
X3
.-
U2 * U3
U3> U2
---+-
ox2
13
Ox3
U12
'*
13
(3
--*-----13
(3
Ox3
Ox2
U13
12
a12
-r 413r. 2 12
3
P3
'2
((/33r, 3
.v1G1 P3
L.
2
(1187)
x2x31!
12
=
t
is even in x>,
Vj
D
i
i---
+ (,b3d, 3]
is odd in x3, and S2r, Ssd involve only integrals of odd functions of
300
CHAP. 11
where q53r,
conditions:
(1188)
(X2 +
2
leads to no shearing stress resultants. Finally, the total normal stress is given by
M3
13
(P3
P2
13
'\
J
(1189)
we obtain
M1 =
where
defines
G1k1J
(1190)
bution due to P3. One can interpret X2, x3 as the coordinates of a point,
called the shear center. The required twist follows from (1190):
k1
(M1 +
M1 + P2x3 P3x2 = the applied moment with respect to the shear center = MT
we can write (a) as
(1191)
k1 =
(1192)
To determine the twist deformation (and the resulting torsional stresses), one must work with the torsional moment with respect to the shear center, not the centroid. For no twist, the applied force must pass through the shear center. In general, the shear center lies on an axis of symmetry. lithe cross section is completely symmetrical, the shear center coincides with the centroid. It is of interest to determine the complementary energy associated with Then torsion-flexure. The only finite stress components are 012, aild V* reduces to
=
dA
(a)
follows directly by substituting (11 89) and using The contribution from the definition equations for 13.
t The total flexural warping function for P3 is
P3
'\
v,P3(
+ 4
SEC. 115.
M2
M2
(11-93)
2. 3.
a pure torsional distribution due to MT the flexural distribution due to F2 the fiexural distribution due to F3
dr.
by
2.
dId,
where the various terms are defined by (1181) and (1187). For example,
r=
(1192):
F2
13
2
F3
13
2
The complementary energy due to pure torsion follows from (1138) and
+
as
C12r
F2
F3_
2 3
r+
JJ
=
+
F2
2FF
F2
dA
(1196)
Jj
1 See Prob. 11il.
2+
3413r, 3)dA
CIA
JJ
302
CHAP. 11
The
coupling term, I/A23, vanishes when the cross section has an axis of
=
+
2)
dA
+ x2)
+ +
(1197)
MT ""F2 Jj 13
F3
12
dA = 0
The remaining terms involve a,a, the shearing stress distribution due to inplane deformation of the cross section, We will not attempt to expand these terms since we are interested primarily in the rigid cross section case. Summarizing, the complementary energy for flexure-torsion with unrestrained warping is given by
1
M2
M2
M2
F2
FF
F2
(11-98)
where
= 2G1k1A5
where the integration is carried out in the X2-X3 sense around S. and
operate on (11 81) and (11
is the
j
(aisd)F2 dS
dS =
2,3
X3
2v1G1P2
dA
(1199)
SEC. 116.
303
In the engineering theory of flex ural shear stress distribution, the cross section
is considered to be rigid, i.e., the distribution due to in-plane deformation is neglected. The consistent continuity condition on the flexural shearing stress
4sajsdSO
definition, the positive sense for
116.
(11100)
We consider the problem of determining the exact shear stress distribution due to F2 for the rectangular cross section shown in Fig. 1115. For con-
rd3
'2
dt3
A = dt
r
Fig. 1115. Notation for rectangular cross section.
1.
Warping functions +
=
an
= =
+
2
304
2.
CHAP. 11
Shearing stresses
012
F2
+
v1G1E2
+
x2x3)
+ xi)]
=T 13
Determination
3) +
of
are
l(d'\2
=0
We can take the solution as
<P2r
.i.
atx3 =
112
X2
= 4)2r
13
12 X2
13
(11101)
012,r
=
0
0j3,r
One can readily show that
F2
(11102)
Determinatio,, of
The boundary conditions for /2d are
2=
1(d2
'\
at
x2 =
atx3 =
Now, the form of (a) suggests that we express q52d
as
f(x2, x3)
(b)
SEC. 116.
305
where
f is an harmonic function. The shearing stresses and boundary conditions expressed in terms of f are
=
v1G1 P2
VLGI F2
2
f 2)
f3=0
It remains to solve V2f = 0 subject to (d). Since the cross section is symmetricaL f must be an even function of x3 and an odd function of x2. We express f as
f=
B0x2
B,, cos
sinh
(2nxx2)
This expansion satisfies V2f = 0 and the boundary condition at x3 = t/2. The remaining boundary condition requires
B0 +
Expanding
B,,
(2nn
co sh
nicd\
c os-
2nicx3
2 x3
< x3 <
(f)
B,, =
(..Y
mr1
mtd cosh
v1G1 P2
.a
1/t\2
t
I
2nirx2
cosh
2nxx2\ 1
nnd
cosh
t
1
/J
[
cJl3,d =
v1G1 F2
sinh
n2
- t
mid
t
13
sin t
2n7tx3
(11103)
[
This system is statically equivalent to zero.
cosh
306
CHAP. 11
To investigate the error involved in assuming the cross section is rigid, we occurs at x2 0: note that the maximum value of
F2
2,
ci 2
Specializing
for x2
0,
(512, d)x20 =
v1G1 F2 d2
13 '1
2nnx3
C,, COS
where
1
2,,
cosh 2,,
Now, C,, decreases rapidly with n. Retaining only the first term in (b) leads to the following error estimate,
4
/
1
I5lz,di
2,.
cosh
Results for a representative range of d/t and isotropic material are listed below.
They show that it is reasonable to neglect the corrective stress system for a rectangular cross section. The error decreases as the section becomes thinner,
i.e., as d/t becomes large with respect to unity:
d/t
2
1
1512,4/512.r
117.
The "exact" solution of the flexure problem involves solving four secondorder partial differential equations. If one assumes the cross section is rigid with respect to in-plane deformation, only two equations have to be solved. Even in this case, solutions can be found for only simple cross sections. When the cross section is irregular, one must resort to a numerical procedure such as finite differences or, alternatively, introduce simplifying assumptions as to the stress distribution. In what follows, we describe the latter approach for a thinwalled cross section. The resulting theory is generally called the engineering
theory of shear stress. We apply the engineering theory to typical cross sections
SEC. 117.
307
and
also illustrate the determination of the shear center and the energy
efficients, 1/Ai (j 2, 3). Figure 1116 shows a segment defined by cutting planes at x1 and x1 + dx1. Since the cross section is thin-walled, it is reasonable to assume that the normal Also, we work stress, Oh, is constant through the thickness and to neglect
qA
xl
Fig. 1116.
thin-wafled segment.
Integrating the axial force-equiwith the shear flow, q, rather than with librium equation, + (a11t) 0
C2S
(11104)
JSA
Equation (11104) is the starting point for the engineering theory of shear over the cross section is known, stress distribution. Once the variation of have shown that the normal stress varies linearly we can evaluate q. Now, we over the cross section when the member is subjected to a constant shear (F2, F3
308
CHAP. 11
constant)
and the end sections can warp freely. Noting that the member is
for this case is
X3
dM2
x2dM3
13 dx1
'2
13
12 dx1
3
X3 +
X2
12
The integrals represent the moment of the segmental area with respect to X2, X3 and are generally denoted by Q2, Q3:
rs
Q2
=j
=
x3tdS = Q2(S,SA)
(11105)
q=
F2
13
F3 -Q2
(11106)
Equation (11106) defines the shear flow distribution for the case of negligible
restraint against warping, i.e., for a linear variation in normal stress. Note
that q is positive when pointing in the + S direction. We consider first the open section shown in Fig. 1117. The end faces are
unstressed, i.e.,
=
F2
=0
F3
q=
Q3
f-Q2
(11-107)
=J5x2tdS
Q2
=$x3tdS
We determine Q2, Q3 and then combine according to (11107). The shearing stress distribution corresponding to F2,
q=
satisfies
ffai2 c/A = dA =
identically. To show this, we expand
=
F2
SEC. 117.
309
Shear center
X2
Centroid
Jjui2dA
J
Equation (b) requires
rsa
J0
dS
=
=0
Now,
Integrating (e) by parts and noting that X2, X3 are principal centroidal axes,t
we obtain
J"
dS =
13
= x2x3t dS = 0 J The shear stress distribution predicted by (a) is statically equivalent to a we evaluate the force F212. To determine the location of its line of moment with respect to a convenient moment center. By applying the same argument, one can show that the shear flow corresponding to F3 is statically
t See Eq. (112).
1112.
310
CHAP. 11
The intersection of the lines of action of the two resultants is the shear center for the cross section (see Fig. 1117).
equivalent to a force,
Example
113
Consider the thin rectangular section shown. We take + S in the + X2 direction. Then, The various terms are + q points in the +X2 direction and q/t =
tx2
j -d(2
F2
q
dx2 =
2
q=
tF2 (d2
F2/'d2
obtained in Sec. 116. Actually, the engineering theory is exact for a rigid cross section, i.e., for r1/E = 0.
x2
Fig. E113
I
Example 114
symmetrical.
We determine the distribution of q corresponding to F2 for the symmetrical section of Fig. Eli 4A. Only two segments, AB and BC, have to be considered since 1Q31 is
Segment AB
=
q
____________________________
SEC. 11-7.
311
Pg. E114A
IC
tw
tf
It H
Segment BC
in Fig. E1I413.
I,
+
[hh1t1 +
q=
xi)]
Note that the actual sense of q is from C to B. The distribution and sense of q are shown
It is of interest to evaluate A2. Specializing (11-96) for a thin-walled section,
if
and substituting for q yields
1
dA
dS
5
2dS
fl
We let
j
2b1t1
l,.
I
3Af\
6A1
2A F
id, l/bf\2
312
CHAP. 11
Fig. E1148
.4
X2)
x3
4.
This factor is quite close to unity. For example, taking as typical, for a wide-flange section,
= 2
(If = 3,
1..
we find
.4f =
k=
0.95
The shearing stress corresponding to F3 varies parabolically in the flanges and is zero in the web. Each flange carries half the shear and
1
61
5
A3
A1 5 hr1
Examp'e 115
Cross-Sectional Properties
This section (Fig. Eli 5A) is symmetrical with respect to X2. The shift in the centroid from the center of the web due to the difference in flange areas is
b2t2
b,t1 + b2t2 +
We neglect the contribution of the web in '2 since it involves
12
(a)
+ ('2)2 =
(b)
SEC. 117.
313
Determination of
Taking S as shown in the sketch, we have
t[(b)2
t2
[,"b2\2
2 X3
(c)
=0
I
I
Centroid
Distribution of q Corresponding to
The shear flow corresponding to F3 is obtained by applying
q=
p3
Q2
and is shown in Fig. El 1SB. The shear stress vanishes in the web and varies parabolically in each flange. Integrating the shear flow over each flange, we obtain
F3
Then, the distribution is statically equivalent to F373 acting at a distance e from the left
flange, where
e
R2d
R
(12)2
12
Since
X2 is an axis of symmetry, the shear center is located at the intersection of R and X2,
314
CHAP. 11
Fig. E11--58
R = F3
q
1'3t2frbf\2
12
X3
R1
The coordinates of the shear center with respect to the centroid are
,c3
= e (1 +
(I
=d
L12
The flexural shear stress distribution is statically equivalent to a torsional moment equal with respect to the centroid. We have defined M1 as the required torsional moment
with respect to the centroid. Then, the moment which must be balanced by torsion is
F3; the required torsional moment with respect to the shear center. Using the approximate theory developed in Sec. 1 1 3, the maximum torsional shear stress in a segment is = M ti
where
.1 =
We consider next the closed cross section shown in Fig. 11 18. We take the origin for S at some arbitrary point and apply (11-106) to the segment Se-S:
q=
=
q1 -1Q3
13
F2
F3
12
(11 108)
Q3 =
where q1 is the shear flow at P. The shear flow distribution is statically indeter-
q1 = con-
SEC. 117.
315
The second and third terms are statically equivalent to F212 and F313. The constant q1 is determined by applying the continuity requirement to the
centerline curve. Since the engineering theory corresponds to assuming the cross section is rigid with respect to in-plane deformation, we use (11100).
Substituting for q.
dS
F2 C
dS
2
(IS
= T (B2
B
Q3)
(B3 Q2)
QdS
QdS
B
(11-110)
27
dS
Each distribution satisfies (11109) identically. Also, the distribution is statically equivalent to a force Fyi,, located Y4 units from the centroid. Note that q = B3 leads only to a torsional moment equal to
f One can interpret (11-109) as requiring the flexural shear stress distribution to lead to no twist deformation, See Prob. 11 14 for the more general expression, which allows for a variable shear
modulus.
CHAP. 11
j = 2,3
(11111)
Substituting for
1
=
'k
2BJQk
(j
k;j, k
2,3)
and
noting that
.1
we
obtain
2dS
42j
dS
QkT
(11112)
Example 116
We illustrate the determination of for the square section of Fig. convenient to take P at the midpoint since the centerline is symmetrical.
Eli 6A.
It is
Fig. E116A
2z
Cross-Sectional Properties
=
=
a2
(a3\
a2
= 4a3t
= (at)(a/2)
5a1
10
it
fdS
= 3.5t
SEC. 117.
317
Determination of
We start at P and work counterclockwise around the centerline. The resulting distribuarc shown in Fig. Eli 6B. Note that + Q2 corresponds tion and actual sense of q due to to a negative i.e., clockwise, q.
Fig. E116B
Evaluation of B3
By definition,
B3=
"
I
dS
it
I
rds.
Using the above results, and noting that the area of a parabola is equal to (2/3) (base) x (height), we obtain
dS
a3
B3 =
F3
F3((
Q2)
4Q2
12
3at
(+ sense clockwise). The two distributions are plotted in Fig. Ell6C. To locate the line of action of the resultant, we sum moments about the midpoint (0 in the sketch:
(M)0 =
/a\
+
4F3 /a\
=
19
aF
e=
19
_
318
CHAP. 11
Fig. Eli 6C
F3/21
q a 1.21
6a
i F3
30
I F3
F3f 4 Q2
Finally, the coordinates of the shear center with respect to the ccntroid (which is A units to the right of 0) are
x2
= e LI = +
16
X3 = 0
Torsional Shear
The shear flow for pure torsion is due to Mr, the torsional moment with respect to the shear center. For this section,
Mr
5L2F3 +
M1
We apply the theory developed in Sec. II 4. One just has to replace M1 with MT in
SEC. 117.
Equation (1161):
Ci = 4at
I
C1
J=
Determination of 1/A3
Applying (11112), we find
1
dS
I-
dS'\
1.276
I) - 3at
We consider next the analysis of a two-cell section and include open segments
for generality. There is one redundant shear flow for each cell. We select a convenient point in each cell and take the shear how at the point as the redundant for the cell. This is illustrated in Fig. 1119: qj represents the shear flow redundant for cell j and the + S sense coincides with the X2-X3 sense to be consistent with the pure-torsion analysis. The + S sense for the open segments is inward from the free edge. For convenience, we drop the CL (centerline)
subscript on S and A.
+5, q
x3
x2
section.
320
CHAP. 11
The total shear flow is the sum of q0, the open cross-section distribution
(q1 =
q2
0),
and
q
q0 +
(11113)
We determine q0 by applying (11107) to the various segments. The redundant shear-flow distribution is the same as for pure torsion (see Fig. 1111). Finally, we obtain a system of equations relating q1, q2 to F2, F3 by applying the continuity requirement to each centerline, t
q--= 0
j = 1,2
(11114)
11115
The shear flows q2, for pure torsion are related by (we multiply (1171) by MT/J and note (1162))
+ a12q2,1 =
a12q1,1 + a22q2,1
2A1
(11--1l6)
2A2
MT
Thus, the complete shear stress analysis involves solving aq = b for three different right-hand sides. The equations developed above can he readily generalized.
Example
117
We determine the flexural shear stress distribution corresponding to F3 for the section shown in Fig. E117A. We locate and P2 at the midpoints to take advantage of
symmetry.
Cross-Sectional Properties
A1 =
A2
202
a2
=
(a3t\
'2
[ a21 + 2[(3at)_4_j
=
4a
Ga
a22
012 =
See Prob. 1!-- 14 for the more general expression, which allows for a variable shear modulus.
_______a _______
SEC. 117.
321
Fig. ElilA
I
We apply
x2
+S,q
2a
a
Distribution of
This system (Fig. E1l7B) is statically equivalent to a moment
2a2(2q1 +
Distribution of q0 Due to F3
--i02 '2
to the various segments starting at points P1. P2. The resulting distribution is shown in Fig. Eli 7C.
Deter,niiuztion of q1.
q2
=
C
.322
q0=+ t
=
1 F3
dS
2F3
7a 70
2 F3 q1 + 4q2 =
Solving (a), we find
q1 =
2F3
161
q2 = +
a 11 F3
322
CHAP. 11
Fig. El 1 7B
q2
2q1a
q2a
q1a
q1
Fig. Eli 7C
P3
14 a
Taking moments about the midpoint of the left web, and letting e be the distance to the action of the resultant, we obtain
M( +
e
+ (3a)
F3) =
eF3
(2
32\
a=
1.61a
(2a
= +O.055a
REFERENCES
1.
2,
WANG, C. T.: Applied Elasticity, McGraw-Hill, New York. 1953. TIMOSHENKO, S. J., and J. N. Oooo!ETt: Theory of Elasticity, 3d ed., McGraw-Hill, New York, 1970.
PROBLEMS
3.
323
Dmt
1952.
HARTOG,
4.
5.
6. 7.
HERRMANN, L. R.: "Elastic Torsional Analysis of Irregular Shapes," I. Eng. Mech. Div., A.S.C.E., December 1965. SOKOLNIKOFF, LS.: Mathematical Theory of Elasticity. McGraw-Hill, New York,
1956.
8. 9.
10. 11.
TIMOSHFNKO, S. J.: Strength of Materials, Part 2, Van Nostrand, New York, 1941. CEiRNICA, J. N.: Strength of Materials, I-Jolt, Rinchart, New York, 1966. DABROWSKI, R.: GekrUmmte diinnwan.dige Trger ("Curved Thin-walled Girders"), Springer-Verlag, Berlin, 1968. KOLLBRUNNER, C. F., and K. Basler: Torsion in Structures, Springer-Verlag, Berlin,
1969.
VLASOV, V. Z.: Thin- Walled Elastic Beams, Israel Program for Scientific Translations, Jerusalem, 1961. ODEN, J. T.: Mechanics of Elastic Structures, McGraw-Hill, New York, 1967.
PROBLEMS
ili. The pure-torsion formulation presented in Sec. 112 considers the cross section to rotate about the centroid, i.e., it takes
U2
= W1X3 = +(01x2
wj =
a1 =
= +w1(x2
wj = k1x1 + c1 a1 = k147
a13 and the
(a)
(b)
Starting with Equation (h), derive the expressions for governing equations for What form do the equations take if we write
+ C2
Do the torsional shearing stress distribution and torsional constant J depend on the center of twist? 112. Show that .1 can be expressed as
Hint:
= = I,,
+
)2
2)
3)2]dA
3)2]dA
0
dA =
Compare this result with the solution for a circular cross section and comment on the relative efficiency of circular vs. noncircular cross section for torsion. 113. Derive the governing differential equation and boundary condition for for the case where the material is orthotropic and the material symmetry axes coincide with the X1, X2, X3 directions.
324
CHAP. 11
114.
=
where
as
1
k1
(a)
is the perpendicular distance from the center of twist to the tangent. One selects a positive sense for S. The sign of p is positive when a rotation about the center of twist results in translation in the +.S direction. We express
lM1
(b)
+ a15
Determine the variation of sections shown.
tf
x3
(c)
of twist
Center
Center of twist
b/2
(a)
b/2
(bi
115.
F2 = ja12 = F3 = dA = x =
for the closed cross section sketched.
116.
= dS = =
dS
0
0
a15 =
Jq
C
(a)
= k =
PROBLEMS
325
Prob. 115
= +
3
x2
Then,
Ct
C +-
Take
Integrating (b) leads to the distribution of Apply (h) to the section shown. = 0 at point P. Discuss the case where a = b.
Prob. 11-6
117.
constant J for the section shown. Specialize for t a. 118. Determine the equations for (j = 1, 2, 3) and J for the section shown. Generalize for a section consisting of "n." cells. I 19. Determine the distribution of torsional shear stress, the torsional constant J, and the distribution of the warping function for the section shown. Take = 0 on the symmetry axis and use the results presented in Prob. 116.
1110. 1111.
satisfy
mA
on S
326
CHAP. 11
Prob. 117
0
+S1
+S2
I
H
Prob. 118
t
I
:f
t.
0
Si
t
t
0
t
0
t
t
a
I
Utilizing the following integration formula,
Prob.119
if (fr,
212,2 +
dS
where 11,12 are arbitrary functions, verify Equation (1196). 1142. Refer to Fig. 1117. Starting with (11107), derive the expressions for the coordinates of the shear center in terms of the cross-sectional parameters.
_a
PROBLEMS
327
11 13. Determine the flexural shear flow distributions due to F2, F3 and locate the shear center for the five thin-walled sections shown.
Prob.
I-
1113
T
d/2
x3
I
(b)
(a)
2/
2/
2/
+
(c)
I I H-a
(dl
-j
I
:L
1
Ca
(el
328
CHAP. 11
1114. We established the expression for the twist deformation (Equation (1131) by requiring the torsional warping function to be continuous. One can also obtain this result by applying the principle of virtual forces to the segment shown as part of the accompanying figure.
Prob.
Arbitrary
closed curve
1114
-44
M'11
44
wi
LsM1
w1 +
dx1
(a)
(b)
2G(2E)
x3
GtE)
G(E)
X21
F
Ic)
2G(2E)
a-
PROBLEMS
329
(J$eT
dA)dx1
Next, we select an arbitrary closed curve, S (part b of figure), and consider the region defined by S and the differential thickness dn. We specialize the virtualstress system such that AG = 0 outside this domain and only is finite inside the domain. Finally, using (11 -51), we can write
dn(&riz)
and Equation (b) reduces to
k1
if
=
The derivations presented in the text arc based on a constant shear modulus G throughout the section, so we replace (d) with
Gk1
G*k1J
Consider a thin-walled section comprising discrete elements having different
material properties. Develop the expressions for the torsional and fiexural shear flow distributions accounting for variable G and E. Determine the
normal stress distribution from the stress-strain relation. Assume a linear variation in extensional strain and evaluate the coefficients of the strain expansion
from the definition equations for F1, Al2, and M3. Apply your formulation
to the section shown in part c of the figure.
INTRODUCTION
1.
2.
There are no surface forces applied to the cylindrical surface. The end cross sections can warp freely.
The warping function consists of a term due to flexure (&j) and a term due to Since is independent of x1, the linear expansion pure torsion
a11 =;+X3 ---Xz 13 '2
F1
M2
M3
+ crj
(122)
and represents the is the pure-torsion distribution (due to flexural distribution (due to We generally determine by applying the engineering theory of shear stress distribution, which assumes that the cross section is rigid with respect to in-plane deformation. Using (121) leads to the
following expression for the flexural shear flow (see (11106)):
(123)
The warping function will depend on x1 if forces are applied to the cylindrical
surface or the ends are restrained with respect to warping. A term due to variable warping must be added to the linear expansion for This leads
to an additional term in the expression for the flexural shear flow. Since (121)
t A linear variation of normal stress is exact for a homogeneous beam. Composite beams (e.g.. a sandwich beam) are treated by assuming a linear variation in extensional strain and obtaining the distributions of from the stress-strain relation. See Probs. 114 and 121. 330
SEC. 122.
FORCE-EQUILIBRIUM EQUATIONS
331
the definition equations for F1, M2, M3 identically, the normal stress correction is self-equilibrating; i.e., it is statically equivalent to zero. Also, the shear flow correction is statically equivalent to only a torsional moment since (123) satisfies the definition equations for F2, F3 identically. In the engineering theory of members, we neglect the effect of variable warping on the normal and shearing stress; i.e., we use the stress distribution predicted by the St. Venant theory, which is based on cons tant warping and no warping restraint at the ends. In what follows, we develop the governing equations for the engineering theory and illustrate the two general solution procedures. This formulation is restricted to the linear geometric case. In the next chapter, we present a more refined theory which accounts for warping restraint, and investigate the error involved in the engineering theory.
satisfies
122.
FORCE.EQUILIBRIUM EQUATIONS
In the engineering theory, we take the stress resultants and couples referred to the centroid as force quantities, and determine the stresses using (121), (123), and the pure-torsional distribution due to MT. To establish the forceequilibrium equations, we consider the differential element shown in Fig. 12--i. The statically equivalent external force and moment vectors per unit
Fdxi/2
.-I.
clxl/2___H
'
dx1
+ dx1
dF+ dx1
4L
dx1
=o
0
dM
dx1
+ m+
x F+) =
332
ENGINEERING THEORY OF
MEMBERS
CHAP. 12
We obtain the scalar equilibrium equations by introducing the component expansions and equating the coefficients of the unit vectors to zero. The resulting system uncouples into four sets of equations that arc associated with stretching, flexure in the X1-X2 plane, flexure in the X1-X3 plane, and twist.
Stretching
dF1
dx1
b1
Flexure in X1-X2
Plane
dF2
dx3 dx1
b2
0
0
m3
+ F2 =
(124)
--s- + b3 =
dx1
dM2
dx1
+ ni2
F3
= 0
Twist
dM1 + m1 = 0
ax
This uncoupling is characteristic only of prismatic members the equilibrium equations for an arbitrary curved member are generally coupled, as we shall show in Chapter 15. The fiexure equilibrium equations can be reduced by solving for the shear force in terms of the bending moment, and then substituting in the remaining equations. We list the results below for future reference.
Flexure in X1 -A'2 Plane
F2
d2it.13
dM3
dx1
ui3
dx1
Flexure
din3
dx1
b2
= 0
(1 25)
in X1 -A'3 Plane
dM2
dx1
in2
d2M2
2 dx1
+ - + b3 = 0 dx1
din2
SEC. 123.
FORCE-DISPLACEMENT RELATIONS
333
Note that the shearing force is known once the bending moment variation is
determined. The statically equivalent external force and moment components acting on the end cross sections are called end forces. We generally use a bar superscript to indicate an end action in this text. Also, we use A, B to denote the negative and positive end points (see Fig. 122) and take the positive sense of an end
x2
MA2
x3
L
Fig. 122. Notation and positive direction for end forces.
to coincide with the corrcsponding coordinate axis. The end forces are related to the stress resultants and couples by
force
=
FAJ = MAJ =
A
(J
1
2 3)
(126)
123.
We
parameters. Applying the equilibrium conditions to a differential element results in a set of six differential equations relating the six force parameters. To complete the formulation, we must select a set of displacement parameters and relate the force and displacement parameters. These equations are generally called force-displacement relations. Since we have six equilibrium equations, we must introduce six displacement parameters in order for the formulation to be consistent. Now, the force parameters are actually the statically equivalent forces and moments acting at the centroid. This suggests that we take as displacement
334
CHAP. 12
parameters
= =
+M
(.128)
Note that (127) corresponds to a linear distribution of displacements over the cross section, whereas the actual distribution is nonlinear, owing to shear deformation. In this approach, we are allowing for an average shear deformation determined such that the energy is invariant.
dx1
dx1
dudXl
dx1 2
2
represents a displacement quantity, and is the external force quantity corresponding to The term dV* is the first-order change in the onedimensional complementary energy density due to increments in the stress resultants and couples.
where
SEC. 123.
FORCE-DISPLACEMENT RELATfONS
335
+ A]\+
dx1
= [AF1u1,1 + /XF2(u2,
+
co3)
+ AF3(u3, 1 +
(02)
+ zXM2a2
j]dx1
(129)
dV* =
3
\CFJ
+ +
=
1
and k1 are one-dimensional measures. Equating (129) and (1210) leads to the following relation between the deformation measures and the displacements:
The quantities
Cl
av*
= u1,
k2
WL
e2
= 112. j . (03
aF2
cM2
23
(02, 1
(12Il)
(123
We see that
1.
2. 3. 4.
e1 is the average extensional strain. e2, e3 are average transverse shear deformations. k1 is a twist deformation. k2, k3 are average bending deformation measures (relative rotations of the cross section about X2, X3).
Once the form of V" is specified, we can evaluate the partial derivatives. In what follows, we suppose that the material is linearly elastic. We allow for the
denotes the initial extensional strain. Now, for unrestrained where torsion-Ilexure is given by (1198). Since we are using the engineering theory
336
ENGINEERING THEORY OF
MEMBERS
CHAP. 12
of shear stress distribution, it is inconsistent to retain terms involving in-plane F1/A, and neglecting deformation, i.e., v1/E. Adding terms due to
= Fie? +
+
where
2AE
+
2GA2
2GA3
LF2
(1212)
+
MT
e?
M1 + F2x3
$5 x2e1 cIA
plementary energy density for the engineering theory. One can interpret
as "weighted" or equivalent initial strain measures. Differentiating (1212) with respect to the stress resultants and couples, and substituting in (1211), we obtain the following force-displacement relations:
e?,
F1
F2
MT
U21
(03
k2
=k2 = k3 +
M2
(1213)
F3
X2 = U3 1 + (02
k3
= (03,
To interpret the coupling between the shear and twist deformations, we note (see Fig. 124) that
U2 U3
X3W1
defines the centroidal displacements due to a rigid body rotation about the
shear center. Comparing (a) with (1213), we see that the cross section twists about the shear center, not the centroid. This result is a consequence of neglecting the in-plane deformation terms in i.e., of using (1212). Instead of working with centroidal quantities (M1, u2, u3), we could have
started with M1. and the translations of the shear center. This presupposes that the cross section rotates about the shear center. We replace u2, u3 (see
Fig. 124) by
U2
+ (01X3
t01X2
U3 =
(1214)
SEC. 123.
FORCE-DtSPLACEMENT
337
where um, U53 denote the translations of the shear center. The terms involving
1+
w3)
+ w2)
(a)
Then, taking
=
F2
a)1,1
(12 15)
US2,
F3
053 1 + (02
with
Since the section twists about the shear center, it is more convenient to work and the translations of the shear center. Once 052, 053, and w1 are
x2
known, we can determine 02, 03 from (1214). We list the uncoupled sets of force-displacement relations below for future reference.
Stretching
F1
e? +
Flexure in X1 -X2 Plane
01:1
F2
GA2
=
=
1 (03
El3
338
CHAP. 12
(1216)
+
I
k2
MT
= Wi,1
The development presented above is restricted to an elastic material. Now, the principle of virtual forces applies for an arbitrary material. Instead of first specializing it for the elastic case, we could have started with its general form
(see (1094)),
[ss
dA]
AP1
(1217)
where represents the actual strain matrix, and denotes a system of statically permissible stresses due to the external force system, AP1. We express the integral as
.
if
A
j=1
and determine
using
taking
+
12
AM2
AM3
13
leads to
e1 =
k2
if
dA
k3
if x2c1 dA if
+ AM1)]dxj =
AP1
Once the extensional strain distribution is known, we can evaluate (b). Using (1218), the one-dimensional principle of virtual forces takes the form
(1219)
The virtual-force system must satisfy the one-dimensional equilibrium equations (124). One should note that (1219) is applicable for an arbitrary material.
When the material is elastic, the bracketed term is equal to dV*, and we can
write it as
dV's
dx1 =
>d1 AP1
(1220)
SEC. 124.
339
[(eq +
AF1 +
AF2
AF3
+
(12-21)
+ El2) AM2 +
At this point, we summarize the governing equations for the linear engineering theory of prismatic members. We list the equations according to the different modes of deformation (stretching. flexure, etc.). The boundary conditions reduce to either a force or the corresponding displacement is prescribed at each end.
Stretching (F1, u1)
F1
I + b1 =
F1
0 (1222)
0,
F1
or u1 prescribed at x1 =
+ b2 = 0 M31 + m3 + F2 =
F2,
F2
2
=
+
U2
i
(1223)
M
u2
(03,1
0,
or F2 prescribed at x1 =
0, L
F3,1 + b3 M2, +
F3
F3
0)3
u3
1+
(1224)
340
CHAP. 12
01T
MT
=
or a1 prescribed at x1 =
0, L
(1225)
m5. =
U2
U3
m1
+ b2x'3 b3g2
X3W1
12-5.
The displacement method involves integrating the governing differential equations and leads to expressions for the force and displacement parameters as functions of x1. When the applied external loads are independent of the displacements, we can integrate the force-equilibrium equations directly and then find the displacements from the force-displacement relations. If the applied load depends on the displacements (e.g., a beam on an elastic foundation), we must first express the equilibrium equations in terms of the displacement parameters. This.problem is more difficult, since it requires solving a differential equation rather than just successive integration. The following examples illustrate the application of the displacement method to a prismatic member.
Example
121
We consider the case where b2 = coast (Fig. This loading will produce flexure in the X1-X2 plane and also twist about the shear center if the shear center does not lie on
the X2 axis. We solve the two uncoupled problems, superimpose the results, and then apply the boundary conditions. Flexure in X1 -A'2 Plane
We start with the force-equilibrium equations,
F21 M3,1 =
F2
F2 =
b2x1
0,
L:
=
With this notation, (c) simplifies to
FAJ
etc.
F2 =
FA2
b2x1
= MAI
XIFA2
SEC. 125.
M3
03, 3
Uz, i
F2
GA2
+ (x1MA3
4
\2E13/
GA
12Ff)
4 \
The general flexual solution (for b2 = const) is given by (e), (f), and (I).
Fig. E121
x2
b3 13
Shear center
Centroid
mr =
Substituting for mr in the governing equations,
=
(01,1
MT
MT = MAT
b2y3x1
U3 =
342
CHAP. 12
Cantilever Case
We suppose that the left end is fixed, and the right end is free. The boundary conditions are
UA2
=
= M53 = MBT
=0
FBZ
=
An
b2L
A3
11. i2
2
MAT =
b2;L
F2 = b2(L
M3 =
=
U2
x1)
Lx1 + x1)
+ b2Lx1
+
+
()
U3 =
b2
4L 4
a)1
h2L
U52
LI
lh2L2
=
=
UB2
13
GL2A2
an illustration, we consider a rectangular cross section and isotropic material with v = 0.3 (d = depth):
As
13
613d2
5
A2
10
By definition, d/L is small with respect to unity for a member element and, therefore, it is
SEC. 125.
343
Fixed-End Case
We consider next the case where both ends are fixed. The boundary conditions are
0.42
= W.43 = (ISA! = 0
=0
FAZ=---2
b2L
M.43 =
MAT =
b2L2
12
b2T3L
M3 =
MT =
u2
U3
b2
IL2
Lx1
+
b2T3
+
L
x1
2)
b2
2L4 + xi)
(u)
= X20)1
(03
b2IL2 = El
=
b2y3
Xj
xi)
xi)
Example 122
We consider a member (Fig. E122) restrained at the left end, and subjected only to forces applied at the right end. We allow for the possibility of support movement at A. The expressions for the translations and rotations at B in tcrms of the end actions at B and support movement at A are called member force-displacement relations. We can
obtain these relations for a prismatic member by direct integration of the force-displacement
t For shear deformation to be significant with respect to bending deformation, G/E must be of the same order as l/A,L2 where A, is the shear area. This is not possible for the isotropic case. However, it may be satisfied for a sandwich beam having a soft core. See Prob. 121.
344
CHAP. 12
relations. In the next section, we illustrate an alternative approach, which utilizes the
x2
FA3
M1, WB 1
UB 3
MA3/
x3
/
= =
Integrating the force-equilibrium equations and applying (a) lead to the following expressions for the stress resultants and couples:
=
MT
(j= 1,2,3)
M52
M2 =
M3
(L
x1)F53
(b)
MB3 + (L
C03,j =
U3,1 =
+ (L xi)F5]
+
= El2
U3 j = 0)2 +
M5.,-
Col,i =
t See Prob. 1211.
SEC. 125.
345
UAI +
L_
(.053
03.43 +
L_ + M83
L2
L2
UA2 + LWA3 +
(052
LX3_ Mjjr +
/L
IL
L3\..
M82
L2
L2_
U83 =
UA3 LWA2
L3\_
co,0
MET
MET =
and write the equations in matrix form:
+ X31B2
US'
L
U82
L3
+
L
L3
GJ
L
L
(082
(053
L2
(f)
+
by fB.
UA2
U.43
LOA2,
(.0,42, (.0A3}
The coefficient matrix is called the member "flexibility" matrix and is generally denoted
We obtain expressions for the end forces in terms of the end displacements by inverting f. The final relations are listed below for future reference:
346
CHAP. 12
=
FB2
+ + +
(co82
WA3)
(coB!
WA!)
(u53 UA3)
+ CA2)
(co81
WAI)
rGJ
MB1
12E
WA1)
L3
(082 UA2) +
+ (0A3)
+ COA2)
+
MB2
UA3)
UA3)
+
+ (2
a2)
WA!)
+ (4 + a2)
M83
j_ WA2
(053
UA2) +
(1)4j)
-1
(4 +
+ (2
12E12
a3)_L_coA3
12E13
a3
where
a2
1+03
introduce the assumption of negligible transverse shear deformation by setting 03 = a3 = 0 The end forces at A and B are related by
We
(j= 1,2,3)
MA! = !V151 MA2
MA3
+ LF53
M53 LF52
M43:
0A3)
WA!)
+ (4 + a2)
MA3
WA2 + (2 02)
T-(uB2 UA2) +
L2
a3)
WA!)
+ (4 + a3)
C0A3
+ (2
SEC. 125.
347
Example 123
We consider next the case where the applied loads depend on the displacements. To simplify the discussion, we suppose the shear center is on the X2 axis and the member is loaded only in the plane. The member will experience only flexure in the X1-X2 plane under these conditions. The governing equations are given by (1223):
F2 =
M3,1
in3
(b)
M3
(03
b2
Once M3 is known, we can, using (b), find F3. Now, we solve (d) for 03 and substitute in (c):
F2
(03,1
F,1
p2,11
Then, M3
E13(u3
b2
and
F2 =
m3
El2 (02
b2,
Finally, we substitute forM3 in (e) and obtain a fourth-order differential equation involving 02 and the load terms:
d4u2
d2 (
b2
'\
(din3
h2)
The problem reduces to solving (i) and satisfying the boundary conditions:
F2 or 02 prescribed or (03 prescribed)
Neglecting transverse shear deformation simplifies the equations somewhat. The resulting equations are (we set 1/GA2 = 0)
(03
= E13(u2, 11
F3 =
d4u3
in3
E13(u2
1
d2
(din3
b2) = 0
'\
348
CHAP. 12
As an illustration, consider the case of linear restraint against translation of the centroid,
e.g., a beam on a linearly elastic foundation. The distributed loading consists of two terms, one due to the applied external loading and the other due to the restraint force.
We write
q ku3
where q denotes the external distributed load and k is the stiffness factor for the restraint.
We suppose rn3 = = 0, k is constant, and transverse shear deformation is negligible. Specializing (k) for this case, we have
(03 = U2
M3 = E13u2,
F2 =
d4u2
F3
E13u2
or u2 prescribed 1
M3 or (03 prescrihedJ
0L
/
where u2
represents the particular solution due to q. Enforcement of the boundary conditions at x 0, L leads to the equations relating the four integration constants. increases with increasing x. The function e_2x decays with increasing x, whereas
For Ax > 0. If the member length L is greater than 2(3/A) = 2Lb (we interpret 3, Lb as the width of the boundary layer), we can approximate the solution by the following:
0
x1
<
Lb:
=
L:
+
+
LB<xl<LLb:
L Lb < x1
u2 =
The constants (C1, C3) are determined from the boundary conditions at x1 =
(C3, C4) from the conditions at x1 = L. Note that C3 and C4 must be of order
and
since
u2isfiniteatx1 =
L.
u2 = M3 =
E13u2,11
= q/k
The complete solution is
U2
(1
cos Ax1)
SEC. 126.
349
q = const
Fig. E123A
//////////// ///////////////////////////////////////// X1
x2
Application 2
The boundary conditions at x1 =
U2,1 =
0
F2 = E13u2,111 = P12
and the solution is
U2 =
PA
Ax1
+ sin Ax1)
Ax + sin Ax)
sin Ax =
= =
e
(1226)
cos Ax =
////////////////////j///
x2
x1
126.
In the force method, we apply the principle of virtual forces to determine the displacement at a point and also to establish the equations relating the force redundants for a statically indeterminate member. We start with the onedimensional form of the principle of virtual forces developed in Sec. 123 (see Equation 1219):
d1 AP1
350
CHAP. 12
Functions
AX
0.000 0.163
0.261
1.4
1.6
1.8
--0.006
0.038
1.8
0.179 0.155 0.128 0.102 0.078 0.056 0.038 0.024 0,012 0.004
0.007 0.002 0.009 0.012 0.014 0.014 0.013 0.012 0.010 0.008
0.007
0.002
where e3,
are the actual one-dimensional deformation measures; d, represents a displacement quantity; AP, is an external virtual force applied in the direction of
The relations between the deformation measures and the internal forces depend on the material properties and the assumed stress expansions. The appropriate
relations for the linear elastic engineering theory are given by (1213). If a displacement is prescribed, the corresponding force is actually a reaction. We use AR,. to denote a prescribed displacement and the corresponding reaction increment, and write (a) as d,. AR,. d, AP, (1227) + AM3)]dx,, where d, represents an unknown displacement quantity. To determine the displacement at some point, say Q, in the direction defined by the unit vector we apply a virtual force APQIq, and generate the necessary
internal forces and reactions required for equilibrium using the one-dimensional
force-equilibrium equations. We express the required virtual-force system as = AP0 (1228) = MJ,QAPQ
ARk==Rk,QAPQ
SEC. 126.
351
leads to
+ k3M1, Q)]dxj
(1229)
This expression is applicable for an arbitrary material, but is restricted to the linear geometric case. Since the only requirement on the virtual force system is that it be statically permissible, one can always work with a statically determinate virtual force system. The expanded form of (1229) for the linearly
elastic case follows from (1221):
dQ
(F2\
+
where
$
+
[(e? +
Fj,Q
i'F3\
+
+
+
M1.
(1230)
M3\
J..J3/
El21
-JM3,Q 1 1dx1
J
if
L
dA
=7JJx2s?dA
Finally, we can express (1229) for the elastic case in terms of V*:
dQ
k
JxLt'AQ
(1231)
ORQ
This form follows from (1220) and applies for an arbitrary elastic material.
Example
124
We consider the channel member shown in Fig. E12--4A. We suppose that the material is linearly elastic and that there is no support movement. We will determine the vertical
Fig. E124A
x2
I)
Centroid
Q
Shear center
352
CHAP. 12
2.
AT =
a1x1
+ a2x1x2 + a3x1x3
M3
(
F2,
System
I
I
We take dQ positive when downward, i.e., in the X, direction. To be consistent, we must apply a unit downward force at Q. The required internal forces follow from
Fig. E124C:
F2,0 = Mr,0
2
1
C
M3,Q =
/L
x1)
\
0
(b)
F5,0 = F3,0
M2,Q
(p. =o
0
(j=
1,2,3)
Fig. E124C
F2,Q
SEC. 126.
353
hzitial Deformations
The initial extensional strain due to the temperature increase is
+ a2x1x, + a3x1x3)
(d)
JJ
= .
Jf
dA = aa3x1
dA = aa2x,
(e)
Deter,ninatio,, of (IQ
Substituting for the forces and initial deformations in (1230), we obtain
dQJ
=
('Lii I
Pc2
c2L
r
5 1))
IL
\)
(f)
cxa,L2
Example 125
When the material is nonlinear, we must use (1229) rather than (1230). To illustrate the nonlinear case, we determine the vertical displacement due to P at the right end Fig. E125
x2
xi
P
"I
Centroid (and
shear
center)
of the member shown in Fig. E125. We suppose that transverse shear deformation is negligible, and take the relation between k3 and M3 as
k3
= a1M3 +
(a)
354
CHAP. 12
Noting that only F2, and M3,0 are finite, and letting e2 = reduces to
L
d0
k3M30dx1
Now,
M3 = P(L
M3,Q = (L
Then,
x1)
x1) x1)
k3 = Pa1(L
Substituting for k3 in (b), we obtain
dQ =
Pa1
P3a3(L
x1)3
+ P3a3
We describe next the application of the principle of virtual forces in the analysis of a statically indeterminate member. We suppose that the member is statically indeterminate to the rth degree. The first step involves selecting r force quantities, Z1, Z2,. , Z.. These quantities may be either internal forces or reactions, and are generally called force redundwns. Using the force-equilibrium equations, we express the internal forces and reactions in terms of the prescribed external forces and the force redundants.
.
=
= M3,0
+
Mf,kZk
(1232)
R1,
kZk
The member corresponding to = Z2 = ''' = Z,. = 0 is conventionally called the primary structure. Note that all the force analyses are carried out on the primary structure. The set (F3, 0, M3, R1, represents the internal forces and reactions for the primary structure due to the prescribed external forces. Also, (F1, k, M3, k, R1, k) represents the forces and reactions for the primary structure due to a unit value of Zk. One must select the force resultants such that the resulting primary structure is stable, Once the force redundants are known, we can find the total forces from (1232). It remains to establish a system of r equations relating the force
redundants. With this objective, we consider the virtual-force system consisting of AZ,, and the corresponding internal forces and reactions,
SEC. 126.
355
that
0,
we obtain
k + kJMi.k)] dx1 =
(1233)
Taking k = 1, 2, ..., r results in a set of r relating the actual deformations. One can interpret these equations as compatibility conditions,
since they represent restrictions on the deformations. To proceed further, we must express the deformations in terms of In what follows, we suppose that the material is linearly elastic. The compatibility conditions for the linearly elastic case are given by
[(eq +
+ +
+
+
dx1
(1234)
(7R
Ic Ic
(k = i,2,..
.,r)
(1235)
1,
2. .
, r)
(1236)
j=
where
= fjk
rr1
=J
+
+
+
+ +
dx1
=>
[(eq + F1.o)F
+
+
+
+
The various terms in (1236) have geometrical significance. Using (1230), is the displacement of the primary structure in the direction of we see that Z1 due to a unit value of ZIc. Since fik = fkJ, it is also equal to the displacement in the direction of Zk due to a unit value of Z3. Generalizing this result, we can write
(1237)
356
CHAP. 12
corresponds to arbitrary points, and i.e., i has the same direction and sense. Equation (1237) is called Maxwell's law of reciprocal deflections, and follows directly from (1230). The term Ak is the actual displacement of the point of application of Zk. minus the displacement of the primary structure in the direction of Zk due to support movement, initial strain, and the prescribed external forces. If we take Zk as an internal force quantity (stress resultant or stress couple), Ak represents a relative displacement (translation or rotation) of adjacent cross sections. One can interpret (1236) as a superposition of the displacements due to the various effects. They are generally called superposition equations in elementary texts.t If the material is physically nonlinear, (1236) are not applicable, and one must start with (1233). The approach is basically the same as for the linear case. However, the final equations will be nonlinear. The following examples illustrate some of the details involved in applying the force method to statically indeterminate prismatic members.
where i, j are
Example
126
This loading (Fig. E126A) will produce flexure in the plane and twist about the shear center; i.e., only F2, M3 and are finite. The member is indeterminate to the first degree. We will take the reaction at B as the force redundant.
Fig. E126A
x2
x2
X3
Shear
center
Primary Structure
One can select the positive sense of the reactions arbitrarily. (See Fig. El 26B.) We work with the twisting moment with respect to the shear center. The reactions are related to the internal forces by
=
R2
Z1
R3
R4
= = +[MT]x,=o
SEC. 126.
357
E1268
x2
x2
R3,d3
R2, d2
ZI = 0
R1
Fig. E126C
__
lB
Mr,o
F2,0
-.p-*'
b
I qe
F2,0 =
x1)
R1,0
R2,0
MTo =
M3,0 =
F1,0
qe(L
q
x1)
= 0 = qL
qL2
x1)2
0
R3,0 =
(b)
F3,0
= M2,0 =
R4,0 = qeL
Bt
F2,1
358
CHAP. 12
F2,1=+1
MT,j =
F1,
e
R2,1 =
1
M3,1=+(Lx1)
=
F3,
= M2, 1 = 0
R41 =
Equation for Z1
We suppose that the member is linearly elastic. Specializing (1236) for this problem,
f 11Z1 =
=
A1
1)2
1)2
)2]dX
(d)
=
and then substituting for the forces and evaluating the resulting integrals, we obtain
fit
Le2
L3
(e)
(L x1)dx1
z1 =
8
Final Forces
The total forces are obtained by superimposing the forces due to the prescribed external system and the redundants:
F2=F20+Z1F2,1 = q(Lx1)+Z1
MT
qe(L x1)
eZ1
M3 =
(L
(g)
= qL
Z1
R3 =
L2
LZ1
e(qL
Z1)
SEC. 126.
359
Example 127
This loading (Fig. E127A) will produce only flexure in the X1-X2 plane. We suppose the material is physically nonlinear and take the expression for k3 as
k3 =
+ a1M3 +
(a)
V.
f // / /
F
Primary Structure
Z1
R7
R3 =
(b)
Fig. E1278
x1
R2, d2
z1 =0
R1,2j
Force System Due to Prescribed External Forces (see Example 126) F2,0 = q(L x1)
M3,0 =
R1,0 =
0
R2,0 = qL
R3,0 =
qL2
360
CHAP. 12
Force
M3,
= L
R1,3=+1
R3,1 = L
R21=1
Compatibility Equation
Since the material is nonlinear, we must use (1233). Neglecting the transverse shear deformation term (e2), the compatibility condition reduces to
J
We substitute for k3 using (a):
dx1 =
J
Now,
(ajM3 +
1W'3
dx1 =
= M3,
JL
+ Z1M3,
x1)2
+ Z1(L xt)
Introducing (g) in (f), we obtain the following cubic equation for Z1:
z?
(asLs) +
+ =
x1)dx1
+
For the physically linear case,
030
and (h) reduces to
fLko(L
xi)dxi]
Example
128
The member shown (Fig. E128A) is fixed at both ends. We consider the case where the material is linearly elastic, and there are no support movements or initial strains. We take the end actions at B referred to the shear Center as the force redundants.
= = Z3 =
Z2
MTB
The forces acting on the primary structure are shown in Fig. E128B.
=
MT,0
P(a
x1)
SEC. 126.
361
Fig. E128A
tP
x2
Shear
Fig. E128B
z3
x3
Fig. E128C
M30
MTO
(
F2,0
P
A
Px3
_________________
362
CHAP. 12
Z = +1
Fig. E128D
Al31
'(
Lx1
'-I
ti
Shear center axis
F2,1
F2,1=+1
M31=Lx1
0
(c)
Z2
= +1
Fig. E128E
M3,2
M72
M3,2 = +1
F22 =
MT,2
z3 = +1
Fig. E128F
Al33
(I
I. x1
////
Shear center axis
=+
F3,
M3,
(e)
SEC. 126.
363
Goinpatibility Equations
The compatibility equations for this problem have the form
JkJZJ =
(k
= 1,2,3)
+ +
fkj =
i:
+
+
Substituting for the various forces and evaluating the resulting integrals lead to the following equations:
[a
[GA2 Pa2
(a3
a2b
2
(L2\ Z1
+
/L\ Z2 =
'\GJJ
El3 \ 3
(g)
GJ
6E13
2b1 PICA
L2GAZ
Z3
Pox3 Z3__
Application
Suppose the member is subjected to the distributed loading shown in Fig. E128G. We can determine the force redundants by substituting for P, a, and b in (h),
P = q dx1 a=
b=L
x1
=
z2
Z3
X3
x1)
6E13
1)
x1q dx1
364
CHAP. 12
where
C=
12E13
L2GA2
q
const in (j),
z2 =
12
Fig. E1280
x2
q(xj)
H
REFERENCES
1.
2.
3.
TIMOSHaNKO, S. J. : Advanced Strength of Materials, Van Nostrand, New York, 1941. HETENY!, M.: Beams on Elastic Foundation, University of Michigan Press, Ann Arbor,
1946.
4. 5.
6.
Elementary Structural Analysis, McGraw-Hill, Noiuus, C. H., and J. B. New York, 1960. ASPLUND, S. 0.: Structural Mechanics: Classical and Matrix Methods, Prentice-Hall,
1966. DEN HARTOG, J. P.: Advanced Strength of Materials, McGraw-Hill, New York, 1952. ODEN, J. T.: Mechanics of Elastic Structures, McGraw-Hill, New York, 1967.
7. 8.
Geac, J. M. and WEAVER, W.: Analysis of Framed Structures, Van Nostrand, 1965. MARTIN, H. C.: Introduction to Matrix Methods of StructuralAnalysis, McGraw-Hill, New York, 1966.
PROBLEMS
12-1. The accompanying sketch shows a sandwich beam consisting of a core and symmetrical face plates. The distribution of normal stress over the depth is determined by assuming a linear variation for the extensional strain:
PROBLEMS
365
= x2k3
M3 =
M3
x2a11 dA
+ Ef13,f)k3
012 A*
f
M3
I
The shearing stress distribution is determined by applying the engineering
is
JJ(aii,i + a21,2 +
ba12 =
Then, substituting for cru,
= (Ek3)x2
(
3)dA = o
(d)
cru, dA M
(Ex2)
(e)
I)equiv
M3,
(d) becomes
a12 =
J'J x2E
dA
366
CHAP. 12
(b)
Apply Equations (e) and (f) to the given section. The flange thickness is small with respect to the core depth for a typical
and G. are
= J)
(c)
dA
=
M3
= (EI)equjv
Refer to Example 121. Specialize Equation (q) for this section and discuss when transverse shear deformation has to be considered. 122. Using the displacement method, determine the complete solution
q = const
F
x1
123.
displacement method. 124. Determine the solution for the cases sketched. Express the solution in terms of the functions defined by (1226).
PROBLEMS
367
Prob. 123
x2
const
Shear
HeH
Prob. 124
(0)
/////////////////////////////7///////////////////
Ib)
Jr
(c)
125. The formulation for the beam on an elastic foundation is based on a continuous distribution of stiffness; i.e., we wrote b2 =
Note that k has units of force/(Iength)2,
ku2
(a)
We can apply it to the system of discrete restraints diagrammed in part a of the accompanying sketch, provided that restraint spacing c is small in
368
CHAP. 12
(k/4E1)'14
c<
Letting k4 denote the discrete stiffness, we determine the equivalent distributed stiffness k from
k=
kd/c
125
>
J
...
r
+
C
r
+
(a)
III
C+C
L,E,11
a/2
/7/7
/7/7
(b)
Consider the beam of part b, supported by cross members which are fixed at their ends. Following the approach outlined above, determine the distribution of force applied to the cross members due to the concentrated load, P.
PROBLEMS
369
a=241t
L=64ft
clit
126. Refer to Example 123. The governing equation for a prismatic beam on a linearly elastic foundation with transverse shear deformation included is obtained by setting b2 = q ku2 in (i). For convenience, we drop
the subscripts:
d4u
We let
&112
k d2u +
u=
d2 7
+
dx
dx 2
uq
Note that
solution is
u
where
a=
b
2(1 2(1
= 0. Determine the expression for the boundary layer length (e3 0). Determine the solution for the loading shown. Assume L large with respect to Lb. The boundary conditions at x 0 are 0) = 0 P
F2
Umax
with
Consider
to vary
126
from 0 to 1.
Prob.
P
/////////)///////////// ////////////)////////
a
370
ENGINEERING THEORY OF
MEMBERS
CHAP, 12
127. Refer to the sketch for Prob. 123. Determine the reaction R and centroidal displacements at x1 L/2 due to a concentrated force Pi2 applied to the web at x1 L/2. Employ the force method. 128. Refer to Example 127. Assuming Equation (h) is solved for Z1, discuss how you would determine the translation u2 at x1 = L/2.
129. Consider the four-span beam shown. Assume linearly elastic behavior, the shear center coincides with the centroid, and planar loading. (a) Compare the following choices for the force redundants with respect to computational effort: 1. reactions at the interior supports 2. bending moments at the interior supports
(b)
to generate influence lines for the redundants due to a concentrated force moving from left to right.
Prob.
129
1210. Consider a linearly elastic member fixed at both ends and subjected to a temperature increase
13
Restrained
of.
a
131.
Prismatic Member
INTRODUCTION
The engineering theory of prismatic members developed in Chapter 12 is based on the assumption that the effect of variable warping of the cross section on the normal and shearing stresses is negligible, i.e., the stress distributions predicted by the St. Venant theory, which is valid only for constant warping
and no warping restraint at the ends, are used. We also assume the cross section is rigid with respect to in-plane deformation. This leads to the result
that the cross section twists about the shear center, a fixed point in the cross section. Torsion and flexure are uncoupled when one works with the torsional moment about the shear center rather than the centroid. The complete set of governing equations for the engineering theory arc summarized in Sec. 124. Variable warping or warping restraint at the ends of the member leads to additional normal and shearing stresses. Since the St. Venant normal stress distribution satisfies the definition equations for F'1, M2, M3 identically, the additional normal stress, must be statically equivalent to zero, i.e., it must
satisfy
dA =
dA =
dA =
(131)
The St. Venant flexural shear flow distribution is obtained by applying the
engineering theory developed in Sec. 117. This distribution is statically equiva-
lent to F2, F3 acting at the shear center. It follows that the additional shear due to warping restraint must be statically equivalent stresses, and to only a torsional moment:
Sfri2 dA
dA
0
(132)
To account for warping restraint, one must modify the torsion relations. We will still assume the cross section is rigid with respect to in-plane deformation.
371
372
En what follows, we develop the governing equations for restrained torsion. We start by introducing displacement expansions and apply the principle of virtual displacements to establish the force parameters and force-equilibrium equations for the geometrically linear case. We discuss next two procedures for establishing the force-displacement relations. The first method is a puredisplacement approach, i.e., it takes the stresses as determined from the strain
(displacement) expansions. The second method is similar to what we employed for the engineering theory. We introduce expansions for the stresses in terms
of the force parameters and apply the principle of virtual forces. This corresponds to a mixed formulation, since we are actually working with expansions
for both displacements and stresses. Solutions of the governing equations for the linear mixed formulation are obtained and applied to thin-walled open and closed cross sections. Finally, we derive the governing equations for geomettrically nonlinear restrained torsion.
132.
JJJaT s d(vol.) =
SSSbT
d(vol.) + JJJJT
d(surface area)
is identically satisfied for arbitrary displacement, L\u, when the stresses (r) are in equilibrium with the applied body (b) and surface (p) forces. We obtain a system of one-dimensional force-equilibrium equations by introducing expansions for the displacements over the cross section in terms of one-dimensional displacement parameters. This leads to force quantities consistent with the displacement parameters chosen. We use the same notation as in Chapters 11, 12. The X1 axis coincides with the centroid; X2, X3 are principal inertia axes; and x3 are the coordinates of the shear center. We assume the cross section is rigid with respect to in-plane deformation, work with the translations of the shear center, and take the displacement expansions (see Fig. 131) as
U1 =
U2
U1
W2X3
W3X2 +
w1(x3
(133)
U3 =
+ w1(x2 x2)
2. 3.
u1, ;2, u53 are the rigid body translations of the cross section. W2, 03 are the rigid body rotations of the cross section about the shear center and the X2, X3 axes.
Note that all seven parameters are functions only of x1. For pure torsion
t See
Sec. 106.
373
the St. Venant theory developed in Chapter 11), one sets f = co1, i = const and For unrestrained variable torsion (i.e., the engineering theory = developed in Chapter 12), one sets f = 0. Since there are seven displacement parameters, application of the principle of virtual displacements will result in seven equilibrium equations.
x3
Shear center
:: e
I
U52
x2
Centroid
(03, 1x2
+f
+ +
2
= =
1(x3 x3)
(134)
+ F2(Au,2, 1 Aw3)
= MR =
quantity Mci, is called the biinornent.
dA
(135)
Note that M1, has units of (force) (length)2 and MR has units of moment. The
t This derivation is restricted to linear geometry. The nonlinear strain expansions are detived
in Sec. 139.
374
To reduce the right-hand side of (a), we refer the transverse loading to the shear center, The additional load terms are
m4,
4picb
=
Then
$5pjcb dA
0, L)
(136)
Au1 + b2 1xu32 + b3 Au,3 mr + rn2 Aw2 + m3 Aw3 + m# Af]dx1 + JF1 Au1 + F2 Au,2 + F3Au,3 + MTLXWI + M2Aw2 + M3 Aw3 +
mj, mj', F,,,
The definitions of
theory. Finally, we equate (b), (c) and require the relation to be satisfied for arbitrary
variations of the displacement parameters. This step involves first integrating (b) by parts to eliminate the derivatives and then equating the coefficients of the displacement parameters. The resulting equilibrium equations and boundary conditions are as follows:
Equilibrium Equations
F1
+ b1 + b2
1
F2, F3,
M2,
F3
0
0
+ b3 =
MTI+m-I-=O
1
+ rn2 =
M3,1 + F2 + m3 = 0 M4, MR + 0
Boundary Conditions at x1
u1
(137)
u1
Us2
US3
or
or or
or
F1 = F2 =
F3
F1
F3
w1=w1
w3=co3
Mr=MT
M2 =
f=J f=J
or or or
Boundary conditions at x1 = L
These are the same as for x1 = For example:
0
or
SEC. 133.
DISPLACEMENT MODEL
375
We recognize the first six equations as the governing equations for the engineering theory. The additional equation,
f=f
O<xj<L
or
end cross section is restrained with respect to warping. If the end cross section is free to warp, the boundary condition is = M4, (+ for x1 = L). To interpret the equation relating MR and the bimoment, we consider the definition for MR, MR = + Integrating (e) by parts leads to
SS4(ail,2 + a13, 3)dA Utilizing the axial stress equilibrium equation,
a12, 2 + c713,
3
is due to warping restraint. Also, we see that one specifies either f or the bimoment at the ends of the member. The condition f = 7' applies when the
+ a11,
we can write
MR
+ JJ'au,j dA
We see that (h) corresponds to the axial equilibrium equations weighted with
respect to
$$(a12,2
+ a13,3 +
i
+ 4(PI
x,,3a13)dS = 0
(i)
MR
In most cases, there is no surface loading on S. i.e., Pi = 0 on the cylindrical boundary. We will discuss the determination of stresses in a later section. We simply point out here that MR involves only the additional shear stresses due correspond to to warping restraint since the St. Venant shearing
FORCE-DISPLACEMENT RELATIONSDISPLACEMENT MODEL
33.
To establish the relation between force parameters and the displacement parameters, we consider (13-4) to define the actual (as well as virtual) strain distribution and apply the stress-strain relations. We also consider the material to be isotropic and suppose there is no initial strain. The stress expansions are
au
a12
1
+
0)3
Gy12
cot, 1(x3
(138)
t M5 =
0 for
and
376
denotes the effective modulus. Although our displacement expansions correspond to plane strain (&2 = = 0), the in-plane stresses vanish on the boundary. Therefore, it seems more reasonable to use the extensional stressstrain relations for plane stress. In what follows, we will take = Young's modulus, F. Consider the expression for The term involving is due to warping of the cross section. This additional stress must satisfy (131), which, in turn, requires 4 to satisfy the following orthogonality conditions :t
where
dA
dA =
dA =
(139)
Assuming (139) is satisfied, and noting that X2, X3 are principal centroidal axes, the expressions for F1, M2, M3, and the reduce to:
EAu1,1
M2
M3
where
E12w2,1 E13a,3 1
j
cIA
1310
We have included the subscript r on E to keep track of the normal stress due to warping restraint. Inverting (1310) and then substituting in the expression for lead to
F1
'Yii
M2
M3
13
(13Il)
and MR expand to
1 (03
i) + fS2
A(u33, i
+ fS3
(1312)
11c01,1
+ +
=
where
Si =
polar moment of inertia =
12
+ 13
x3q5,2)dA
=
t F1 =
M2 =
SEC. 133.
DISPLACEMENT MODEL
377
+G
The essential step is the selection of which, to this point,must satisfy only the ortliogonality conditions (139). To gain some insight as to a suitable form let us reexamine the St. Venant theory of unrestrained torsion. We suppose the section twists about an arbitrary point instead .of about the centroid as in Sec. 112. The displacement expansions are
for
u2 = w1(x3
U3
coj(x2
=
where
i=
=0 a12 =
2
(x3 x3)]
+
follow from the axial equilibrium
The equation and boundary condition for equation and boundary condition,
mA
on S
We can express
as
=
for tile stresses become
x'2x3
a12 =
M1
cr13
3
x3)
-
x2)
Since
depends only on the cross section, it follows that the stress distribution
378
torsional constant are independent of the center of twist. Also, one can showt that = 0 SJcbt,2dA = 2)dA 3)2]dA 3 $S[(4)t. 2) +
and
Suppose we take 4' = The constants (C, 4' to satisfy (139), and we obtain
C=
= =
dA
Now, one can shows that the equations for are identical to the equations for the coordinates of the shear center when the cross section is considered to be rigid with respect to in-plane deformation. That is, the warping function for unrestrained torsion about the shear center is orthogonal with respect to
1, x2, x3.
C T3x2
+ x2x3 +
(1314)
is a permissible warping function. The cross-sectional properties and forcedisplacement relations corresponding to this choice for 4' are listed below:
Properties
S2 =
14,
=
14,
'1315
2)
3)2]dA
Shear Stresses
a12 =
F2
+ G(x3co1,1
+
(1316)
=
t See Sec. 112 and Prob. 112. See Prob. 131.
+ G(x2w1,1 + f4)t,
SEC. 134.
379
Force-Displacement Relations
MT = MR =
G11w1, i
w1, us2, j
W3
+ .T2F3
w1,
(1317)
F3
j
+ W2 + x2(f
Wi,
We
setting Er
to MR =
0. Then, M4, 0, and thc seventh equilibrium equation reduces Specializing (1317) for this case, we obtain
f to1,
(x3F2
GJW1,1
x21'3)
(1318)
MT =
and
F2(1
\J1
F3( x2x3
F3!'! + -- +
(1319)
14,
us3,1 = to2 +
F2(
G\
The shearing stress distributions due to F2, F3 do not satisfy the stress boundary condition 0 on S +
=0
for arbitrary F2, F3. Equations (1319) are similar in form to the results obtained in Chapter 12, which were based on shear stress expansions satisfying (a) identically on the boundary.
Finally, we point out that torsion and flexure are uncoupled only when
warping restraint is neglected (F. = 0). Equations (1317) show that restrained torsion results in translation of the shear center. We will return to this point in the next section.
134.
To obtain an indication of the effect of warping restraint, we apply the theory developed in the previous section to a cantilever member having a
rectangular cross section. (See Fig. 132). The left end (x1 =
0)
is fixed with
380
x3
T
2b
Sect. A-A
to both rotation and warping while the right end (x1 = L) is free to warp. The boundary conditions are
respect
x1=O
x1=L
M1=M
(a)
M1,1+m1=O
Force-Displacement Relations (See (1310) and (1312). F2 = F3 = 0)
(b)
(c)
Note that
= M1 = =
G11w11
+ +
(d)
At xj= 0,
(e)
Atx1
= M
1,1=0
Xj
We start with (b). Integrating (b) and enforcing the boundary condition at = L leads to (1320) M1 = M
SEC. 134.
381
G11co11 + i +
=M
=
t
(f)
(g)
G111,
and then substituting in (g) lead to
(h)
(i)
where 2 is defined as
G
[i,,
13
- 21
Note that has units of(1/length)2. The solution of(i) and (h) which satisfies the boundary conditions (e) is (we drop the subscript on x for convenience)
f=
= =
{l
cosh
+ tanh
[sinh
sinh
+ (1 cosh
tL]}
(1322)
+ '1 The rate of decay of the exponential terms depends on we can take tanh )L 1, and the solution reduces to
I,fr
For
)L >
2.5,
As
M
(J)
We see that is a measure of the length, Lh, of the interval in which warping restraint is significant. We refer to Lb as the characteristic length or boundary layer. By definition,
0
(1324) (1325)
The results obtained show that is the key parameter. Now, depends on the assumed warping function. If the ratio G/ET and on terms derived from
382
we take 4)
(1326)
At this point, we restrict the discussion to a rectangular section (see Fig. 132) We evaluate the various integrals defined by (1315) and write and 4) = the results as
J=
= K1a3b =
(1327)
where the K's are dimensionless functions of b/a. With these definitions, the expression for takes the form
1 /G'\112 2= --- K2
is essentially
constant. Assuming E 2.6G and K1 3.2, we find 2/b and Lb 2b. The influence of warping restraint is confined to a region of the order of the depth. Although this result was derived for a rectangular cross section, we will show later that it is typical of solid and also thin-walled closed cross sections.
Table
b
/<4
131
2 3
10
.156 .450
.683 .964
We consider next the problem of locating the center of twist. We utilize the and large solution corresponding to 4)
M1
fC =
,,
(1329)
e
reduces to
SEC. 135.
MIXEIJ FORMULATION
383
Us31
= x3(f = x2(f
(1330)
By definition, the translations are zero at the center of twist. Setting i12 =
see that the center of twist approaches the shear center as x increases. The maximum difference occurs at x = 0 and the minimum at x = L.
We
1
2L!1
0,
oc,
and g =
1.
We first review briefly the basic variational principles for the three-dimensional formulation. The principle of virtual displacements requires
Au d(surface area)
to be satisfied for arbitrary Au and leads to the stress-equilibrium equations and stress-boundary force relations. Note that g is a function of Au and is obtained using the strain-displacement relations, The stress-strain relations can be represented as
since, by definition of the complementary energy density,
=
By
av*
combining (a) and (b), we obtain a variational principle which leads to both sets of equations. The stationary requirement,
liT11
V*)d(vol.)
d(surface area)] =
E(u),
(1333)
and
b prescribed, is
See Ref. 11 and Prob. 1028. Reissner's principle applies for arbitrary geometry and elastic
material. This discussion is restricted to linear geometry. The nonlinear case is treated in Sec. 139.
384
The essential point to recognize is that Reissner's principle allows one to work with and u as independent quantities. In a displacement formulation (Sec. 133), we take as a function of u, using the stress-displacement relations V* reduces to V, the strain-energy density. In a = and = mixed formulation we start by introducing expansions for the displacements. The Euler equations for the displacement parameters are obtained by expanding (a). This step leads to the definition of force parameters and forceequilibrium equations. We then generate expansions for the stresses in terms
[f$(CT
V*)dA]dxj =
(1334)
The first step was carried out in Sec. 132 and the expanded form of 5$
dA
is given by (b) of Sec. 132. Letting represent the complementary energy per unit length along X1, and using (134), the stationary requirement on the stresses (Equation 1334) expands to
+ +
1 W3) + F3(u,3, i
+ co2) + +
=o
1335)
In order to proceed further, we must express in terms of the force parameters (F1, F2, ..., MR). Equating the coefficients of each force variation to zero results in the force-displacement relations. Instead of applying (1334), one can also obtain (1335) by applying the principle of virtual forces to a differential element. We followed this approach in Chapter 12 and, since it is of interest, we outline the additional steps required for restrained torsion. One starts with (see Fig. 133)
5V* dx1 =
op
+ [$$uT ISp
(a)
The boundary forces are the stress components acting on the end faces. Taking u according to (133) and considering only MR. we have
$5
oiiTudA = SJISrsTudA
= (ISMTWI + ISM4f)
where the plus sign applies for a positive face. The virtual-force system must be statically permissible, i.e., it must satisfy the one-dimensional equilibrium equations. This requires const
=
Then,
dxi{f,i
+ w1,1
1 ISMT}
(d)
SEC. 135.
MIXED FORMULATION
385
The first procedure (based on (1334)) is more convenient since it avoids introducing the equilibrium equations. However, one has to have the straindisplacement relations. In certain cases, e.g., a curved member, it is relatively
aM,5
6Mg, +
WI
o,1+w11dx1
f + f,1 dxj
Fig. 133. Virtual force system.
to establish the force-equilibrium equations by applying the equilibrium conditions to a differential element. We obtain the force-displacement relations by applying the second procedure (principle of virtual forces) without having to introduce strain expansions.t In what follows, we consider the material to be homogeneous, linearly elastic
easy
and isotropic. To simplify the treatment, we also suppose there is no initial strain. The complementary energy density is
=
if
dA +
if
(13-36)
It remains to introduce expansions for the stress components in terms of the force parameters such that the definition equations for the force parameters
are identically satisfied. Considering first the normal stress, we can
F1
M2
M3
(a)
where
JJx2b c/A
c/A = 0
(b)
Note that we have imposed a restriction on q5. The complementary energy due to c11 expands to
,
1
t The approach based on the principle of virtual forces is not applicable for the geometrically
nonlinear case. See 1). Problem F1 = = M3 = 0 for
386
in (1335), we obtain
F1
W3, I
L.13
f,
These expansions coincide with the corresponding relations obtained with the
displacement model (see (1310)).
The shearing stress distribution must satisfy thc definition equations for F2, F3, MT and MR identically. We can obtain suitable expansions by adding
a term due to warping restraint to the results for unrestrained torsion and
fiexure. We write
is the unrestrained is torsion distribution; and is the distribution due to restrained torsion. Since we are assuming no in-plane deformation, the fiexural distribution for a thin-walled section can he obtained by applying the engineering theory
where
(1337)
developed in Sec. 117. For a solid section, we utilize the results of Sec. 115,
taking v = 0. The shear stress distribution for unrestrained torsion is treated in Sees. 112 through 114. Since the restrained-torsion distribution is statically equivalent to a torsional moment, we have to distinguish between the unrestrained and restrained torsional moments: MT =
+
(1338)
= =
It remains to determine We follow the same approach as in the engineering theory of flexural shear stress, i.e., we utilize the axial equilibrium equations and stress boundary condition:
('12,2 +
mA
0
on S
Differentiating the expression for au and noting the equilibrium equations, we obtain
('11,1 =
F2
13
F3
12
MR
Since
0,
it
follows that
due to MR:
y r ('12.2 + ('13.3 =
MR
4
0
(mA)
-
(on 5)
SEC. 135.
MIXED FORMULAflON
387
dA =
dA =
from
(13 40)
= J$[(x3
Noting (1340), we see that
(c)
=
where
(1341)
MT = = =
We take
flow qr
at a free edge
and
=
qV
11
0
(1343)
at a free edge
ensure that
(1344)
=
=
(IS 0
and equating to (1341). by evaluating Finally, we determine We consider next the complementary energy density. We write the expanded form of the shear contribution as
V*,hear
JJ
We
(134)
in Sec. 115. For convenience, these have evaluated and results are summarized below (See Equation 1198) /z'2 I 13 I. 112 Vf = + +
23
3
(a)
LI
2GJ
11*
V uf
f See Prob. 132.
388
The coupling term, 1/A23, vanishes when the section has an axis of symmetry.
= 0 is a consequence of our assuming the cross section is rigid with respect to in-plane deformation. We evaluate using (1339) ((1343) for the thin-walled case), and write the results as
Also
if if
(1346)
where Cr is a dimensionless factor which depends on q5. The coupling between unrestrained and restrained torsion is expressed as
(13-47)
It is obvious that = 0 for a thin-walled open section since is an odd function of n whereas a' is constant over the, thickness. We will show later that it is possible to make vanish for a closed section by specializing the homogeneous solution of (1343). Therefore, in what follows, we will take
Cur
Finally, we write the coupling bctwcen flex ural and restrained torsion as
+
(1348)
1
+ X2rF3Mg)
= 0 since is symmetrical and a' is antisymmetrical with respect to the X2 axis. We substitute for in (1335). replace Mr with + Ccj,MR, and equate
the coefficients of oF2, OF3,
relations are
U521 0)3
"F2
+ -- +
A3
F3
x3,
Us3,1 + 0)2
= GA23
=
C,
(1349)
Wi.j
I+I =
MR +
(x3,F2 + x2,F3)
The corresponding relations for the displacement model are given by (1312). Up to this point, we have required to satisfy the orthogonality relations and also determined a' such that there is no energy coupling between au and (C,1, = 0). If, in addition, we take ISC If" X3X2 -t-- X2X3 -r 5)
SEC. 136.
389
Ct, = +1 MrT=+MR
(1350
Note that
is the warping function for unrestrained torsion about the shear center. We discuss the determination of 4> in Secs. 137 and 138. One neglects shear deformations due to flexure by setting
(13-51)
Similarly, we neglect shear deformation due to restrained torsion by setting Cr = X2r = X3r
1 =
0
(1352)
This assumption leads to the center of twist coinciding with the shear center and
(1353)
AK:'
In what follows, we outline the solution procedure for restrained torsion and list results for various loadings. We then discuss the application to open and closed cross section.
136.
We suppose only torsional loading is applied. The force-displacement relations are obtained by setting F2, F3, co2, w3 equal to zero and C,1, +1 in (1349). For convenience, we summarize the governing equations below.
Equilibrium Equations
MT. 1
+ 112T = 0
MT
Force-Displacement Relations (4>
E,.l4,j
1
AK
= GJw1,1
GJ
+f)
will be positive.
390
Boundary Conditions
MT
or or
=
dx1 =
+
(1354)
Substituting (c) in (b) and (1354) leads to the governing equations for w1 and f:
(1 + Cr)O)i,t
GJ(w1,1 + f) = 0
After some manipulation, (f) becomes
Eric,,
C1x1
(C1
cc =
1
+ Cr
ErI#
(1355)
22
Equation (g) corresponds to (h), (i) of Sec. 134. The general solution for f and has the following form:
f
Wi
C3 cosh 2x + C4 sinh 2x
+
(1356)
+ C2 +
sinh 2x + C4 cosh 2x)
SEC. 136.
391
The significance of A has been discussed in Sec. 134. We should expect, on the basis of the results obtained there, that AL will be large with respect to unity for a closed section. We will return to the evaluation of A in the next section. In the examples below, we list for future reference the solution for various loading and boundary conditions.
Example
131
CantileverConcentrated Moment
Fig.
E131
X1
x=O
x=L
Starting with (1354), we set
cv1=f=O
f. =
0
mined from
w1f0
atx=0 atx=L
M[
GJ
WI
coshA(Lx)
cosh AL
{sinh AL sinh A(L x)}
(1357)
[x
M
r
LA
[
M
t The corresponding solution based on the displacement model is given by (1322), (1326). The expressions for! differ by a minus sign. This is due to our choice of We took in the displacement model and = in the mixed model.
_______
392
Note that C, 1 when the complementary energy term due to the restrained torsion shear stress (o') is neglected. The translations of the shear center are obtained by integrating
u,2,, =
and requiring u,2, 0s3 to
=
0.
vanish at x =
U,2 =
X3.U
u=j
Let
M1dx=xco1
(13-58)
u,2
=u,3 +
Substituting for and w1, we obtain t
X2
x3) = 0 X2)=0
gx3,
+
x
(1359)
x1]
The limiting
of g occur at x =
(1360)
1
k) is an axis of symmetry for the cross section. Also, x2. = = 0 if we neglect shear deformation due to the restrained shear stress and, in this case, the center of twist coincides with the shear center throughout the length.
Note that Xi,. = 0 if X,, (j
x3,.
Example 132
We consider next the case where warping is restrained at both ends; the left end (x =
0)
is fixed and the right end rotates a specified amount w under the action of a torsional moment. The boundary conditions are
x=0 x=L
co1=f=0
ce1=w
[=0
f See (1331), (1332) for the displacement model solution. There is no twist or translation at x = 0. We determine g(O) by applying L'I-IOspital's rule
to (1359).
SEC. 136.
393
To simplify the analysis, we suppose there is no distributed load. Starting with the
general solution,
MT = C1
f = C3 cosh Ax + C4 sinh
=
+ C2
C3=
=
C1L
11c
s = sinh AL
GJ = cosh AL
4
C, [2(1 c)11
(0
1
C1(
=
+
Ax
/1c'\
+
(1.
sinh Ax
(1361)
(01
= ci {i =
M1,
[cosh Ax +
sirih Ax]}
= ErI#A
{sinh Ax +
(L__f) cosh
Ax}
We write the relation between the end rotation, (0, and the end moment M, as
M
=
where L,ff denotes the effective length:
L,5
=L
2C
(c 1'\1
(13-62)
= L(1 cC3)
The following table shows the variation of with AL. For AL > 4, C2 2/AL. Note = 1 if transverse shear deformation due to restrained torsion is neglected. that
AL
C3
0.5
1
0.98 .924
.76
.60
.48
394
Example 133
MT =
C1 ?flX
f= coshAx +
C1
C3
C4.
C1
flIX
x + C2
m /x2
(a)
(C3
We consider the boundary conditions to be identical at both ends and measure x from the midpoint (Fig. E133). Symmetry requires
MT=O}
and (a) reduces to
atx=O
(b)
MT =
[=
sinh Ax +
(13-63)
= C2
cosh xx
We treat first the case where the end section is fixed with respect to both rotation and
warping. Requiring (1363) to satisfy
f=
results in
=0
sinh Ax}
at x = L/2
(a)
I=
CO1
{x
mL2 fi[
U
(
/x"t21
j+
C )
(cosh ,,x
c)
MT = mx
= ,nL1
1
(1364)
+
AL
c=
AL
s=
AL
The solution represents an upper bound. A lower bound is obtained by allowing the
section to wrap, i.e., by taking
SEC. 137.
395
Fig. E133
x1
H
f=
{x
xc
sinh
mL2 (1 [
j+
C,
c)
mx
(
x+fsinhAx
(1365)
IC,
I
AL
cosh Ax
c=
cosh
137.
In what follows, we apply the mixed formulation theory to a wide flange section and also to a channel section. We first determine the cross-sectional and then obtain general expressions for properties corresponding to = the stresses in terms of dimensionless geometric parameters. Before discussing
the individual sections, we briefly outline the procedure for an arbitrary section.
Consider the arbitrary segment shown in Fig. 134. We select a positive sense for S and an arbitrary origin (point P). The unrestrained torsion warping function is obtained by applying (1129) to the centerline curve and requiring the section to rotate about the shear center.t
is positive when translation in the + S direction rotates the position vector about the + X1 direction. The unrestrained torsional shear flow is zero
By definition, Ic1 = We work with q" rather than and mixed sections where one generates q" in terms of
396
sP
(1367)
Note that one can select the sense of S arbitrarily. Also, varies linearly with S when the segment is straight. The constant is evaluated by enforcing the * F1 = 0), orthogonality condition
dS
=0
If the section has an axis of symmetry, = 0, if we take P on the symmetry ?v.t2 = M3 axis. The remaining orthogonality conditions (a'j1 0),
$4x2tdS =
Shear center
IPsc
x2
When the section has branches, we apply (1367) to each branch. One has
only to require continuity of 4) at the junction point. As an illustration, consider the section shown in Fig. 135. The distribution of 4) for the three branches is given by A B 4)p+$gpscdS 4)
BC BD
b4)B+JopscdS
C
and B D.
SEC. 137.
397
=
145
(1368)
+ J
q5tdS =
(1369)
We start at a free edge and work inward. A +q points in the +S direction corresponds to _qr, i.e. qr acting in the S (see Fig. 135). Then, a + direction, If the section has an axis of symmetry, is an odd function with
respect to the axis and
x3
is an even function.
S.q
p
S,q
Once
and
= JJqYdA = 542tdS
Cr =
if
In order to evaluate XZr, X3r, we need the fiexural shear stress distributions. and write We let q(J) be the distribution due to
qU)
(1371)
j=2 j=3
'(=3
k= 2
398
q q
X3r +
X21.
J I
X3r
3dS
(13-72)
If X2 is an axis of symmetry,
and X3r = 0. By analogy, x2, = 0 if X3 is an axis of symmetry. apply for an arbitrary thinThe definition equations for Cr, Ia,, x2r, and
walled section. When the section is closed, we have only to modify the equaand We will discuss this further in the next section. tions for 4) Example
134
Symmetrical I Section
The I section shown (Fig. El 34A) has two axes of symmetry; it follows that the shear center coincides with the centroid and the warping function is odd with respect to K2, X3.
Fig. E134A
x3
1,
Applying (1367), we obtain
q5=O
q5 =
S
forweb
for flange
SEC. 137.
399
The shear flow vanishes at S = b/2. Applying (1369) and starting from pt. A, we find
=
The distributions of
=
S-b/2
(25)2]
and q' are shown in Fig. E134B, where the arrows indicate the
Plot
Plot of qr
= b/h
3 =
ht3
th5
(t)2
8(1 +
=
The dimensionless parameters occurring in the solution of the differential equations for the mixed formulations are and AL (see (1355)). Using (c) and assuming a value of 1/3 for Poisson's ratio, we write
[3(1 +
=
AL
400
The coefficients
0.75
2.4 2.66
3.2
4.22
0.50
6.93
1. The warping parameter, ),L, depends Since (t/h)2 << 1 and 0(1), we see that on t//i as well as L/h. This is the essential difference between open and closed cross sections. For the solid section, we found that AL = 0(L/h) and, since L/h is generally large in com-
parison to unity, the influence of restrained warping is Iocalized.f The value of AL for an open section is O(L//,) 00/1,) and the effect of warping restraint is no longer confined to a region on the order of the depth at the end but extends further into the interior.
We consider next the determination of the stresses due to restrained warping. The
general expressions are
M4,
dTts
7
6
Using the distribution for b and qr shown above, the maximum values of normal and shear
stress are
=
To gain some insight as to the relative magnitude of the various strcsses, we consider a member fully restrained at one end and subjected to a torsional moment M at the other end. This problem is solved in Example 131. The maximum values of the moments are
tanh
AL
atx = 0
J
= C5M
We substitute for the moments in (f), (g) and write the results in terms
the maximum
;.L
SEC. 137.
shear
tanh
(f))
Mt
0.75 0.50
2.11 2.31
1.5 1.67 2
Since
and
0(d)
The additional shearing stress (at) is small in comparison to the unrestrained valu
Therefore, it is reasonable to neglect the terms in the complementary energy density due to ic., to take C, = 1 for an open section. We will show in the next section 0 and that this assumption is not valid for a closed section.
Example
135
Channel Section
symmetry,
We consider next the channel section shown in Fig. El3SA. Since X2 is an axis of = x3, = 0. The expressions for the location of the centroid, shear center,
Fig. E135A
S
Shear
center
x2
402
and 12 are
I+
I+
e=b
=be
th3
b h
1.00
0.75
0.50
We determine by applying (1367) to the three segments. Taking S as indicated above, and noting that is odd with respect to X2, we obtain:
Segment 12
6
Psc
=
Segment 23
hh(
- -S
bh(
2S\
The distribution is plotted in Fig. E135B. Since? < 1/2, the maximum value of q5 occurs
at point I (and 4). We generate next the distribution of and using (b):
at that point)
Segment 12
S
bin
152
Segment 23
/
=
The distribution of
'\
s2
SEC. 137.
403
Fig. E135B
(1 e)
Distribution of
Fig. E13SC
D2
D1
)+Mi
D2
I
D2
0
Distribution of qr/1242t
I=
(1
+
I h5
c
(t'\2
=
(t\2 f=
Cs =
AL
(t)
404
and
ratio equal to 1/3. Note that the comments made for the wide-flange section also apply
to the channel section.
c =
h
1
2.33
2.65 3.4
2.55
3.39
0.75 0.50
5.24
In order to evaluate X2r, we need the flexural shear stress distribution due to F3. Applying (11106) leads to
Segment 12
4(3)
Segment 23
4(3) =
S)
The distribution is plotted in Fig. E13SD; the arrows indicate the sense of q for a +F3.
Fig. E135D
tF3
Distribution of
/2
lb/it
SEC. 138.
405
Substituting for
=
=
(t\2
(1 +
0.926
1.03
0.5
In Example 131, we determined expressions for the coordinates of the center of twist in terms of .'c,, and It is of interest to evaluate these expressions for this cross section.
The coordinates at x
= X2
i1
Substituting for
and evaluating
we obtain
X2
=
=
0.5
0.476 0.625
0.836 0.485
138.
We treat first a single closed cell and then generalize the procedure for multicell sections. Consider the section shown in Fig. 136. The +S direction is from X2 toward X3 (corresponding to a rotation about the +X1 direction). Using the results developed in Sec. 114, the shear flow for unrestrained tor-
sion is
q =-1-C
2A
406
where A is the area enclosed by the centerline curve. The shearing stress varies
C'\
) =
x2
=
We determine
by enforcing
dS C
(1373)
are identically satisfied by definition of the shear center. t The shear flow due to is defined by (1369),
q=-7--q 14
+ Q4
t Noting that x2t =
dQs/ds, we can write
#x24.t = We merely have to identify this term as the moment of the flexural shear stress about the shear center. See Prob. 11-12.
SEC. 138.
407
where
is
between qU and
(1374)
Noting that
f
=
dS
(1375)
The flexural shear flow distributions for F2, F3 are generated with (11110). We merely point out here that there is no energy coupling between qU and
quqf
0
(f)
tion, i.e., w1
One can interpret (1374) and (f) as requiring qr to lead to no twist deforma0. We have expressed the fiexural shear flows as (see (1371)):
ft qjij=q
(J)
FJ_0)
j=3 k=2
Cr
Finally, the definition equations for the cross-sectional properties have the same form as for the open-section:
Eq. 1370
Eq. 1372
X2 is an axis of symmetry. Then, is an odd function of x3. If we take the origin for S (point p) on the X2 axis, = 0. Also, is an even function of x3 and = 0. In what follows, we illustrate the application of the procedure to a rectangular cross section.
Example
136
ci + b
fa b\
when a =
b,
i.e., a square
408
T
Centroid
- x2
2a
Fig. E136B
(ab
+
b
Distribution of
= at
b'\ S2
(
+ bJ 2
(a b\
for segment 12
for segment 23
= (Q4 +
and evaluate
with (1375):
/ a+b1\
dS
.11
ab
SEC. 138.
409
The distribution of
j(
2a\)
2a(S
1(S\2\
j/
2a
2 \a+b
corresponds to q' acting in the clockwise and is plotted in Fig. E136C. Note that ( S) direction for + Ms-. Also, D is negative for b > a.
Fig. E136C
x3
T
2a
b
q'/D
qr(+
D-
410
J=
=
16a3t
(neglecting
4a5t [2(1
4
LATh
+
+
+
5(1
9(1
I
x2, = x3,.
(G'\ /
0
\) 1/2 L
and = -a
L;f for
.'
3 8
2
3
cc 10.43 4.41
0.98
1.27
0.0877 0.185
1.39
We found
(g)
=0
T)
IL
for an open section. Our results for the single cell indicate that
=0
C,>> 1
C,
1
for a closed section. We obtained a similar result for using the displacement-model formulation for a solid section. Since is due to the restrained shearing stress (q'), we see that shear deformation due to q' cannot be neglected for a closed cross section. We discuss next the determination of the normal and shearing stresses due to warping. The general expressions are
isq
q'
t
tie
SEC. 138.
411
The maximum normal stress occurs at point 2 while the maximum shear stress can occur
at either points I or 3. We consider the same problem as was treated in Example 134, i.e., a member fully restrained at one end and subjected to a torsional moment M at the other end. We express the stresses in terms of ag,, the maximum shear stress for unrestrained torsion,
M(
= J
which reduces to
MC
=
since
M =
7
=
=
tanh 2.L
0$
,nax,I
[3C,1112
S.C
The variation of
and
=
1
h/a
c,
(point 2)
0
1.04
(point 1)
0 0.35 0.46
(point 3)
0
2
3
1.51
+0.44 +0.65
For large tanh I and we see that both the normal and shear stress are of the order of the unrestrained-torsion stress. In the open section case, we found the restrainedtorsion shear stress to be of the order of (thickness/depth) times the unrestrained shear
stress.
We nttmber the cells consecutively and take the +S sense from X2 to X3 for the closed segments and inward for the open segments. The total shear
flow is obtained by superimposing the individual ccli flows
q' =
qU =
We let
WIT
(U
412
e
"2
q,S
q1 ,S1
The constants C1, C2 are determined by requiring each cell to have the same
twist deformation, w1,
Enforcing (1167),t
=
for each cell leads to
2A
f
= a21
,Jsj t
dS
dS
A = {A1, A2}
=
= Psc
(1376)
We start at point P1 in cell 1 and integrate around the centerline, enforcing continuity of 4, at the junction points b, c, and d. For example, at b, we require
SEC. 138.
413
and 4),,,:
Jb
= 4)e +
Psc
dS =
dS
Repeating for points C and d results in the distribution of 4) expressed in One can easily verify that 4) is continuous, i.e., & determined terms of determined from segment cdcL. Finally, we from segment ca is equal to
evaluate
by enforcing
JJ4)dA=J4)tdS=O
where the integral extends over the total centerline. Note that
is taken on an axis of symmetry. The shear flow for restrained torsion is obtained with (1369):
a
as
0 if P1
The steps are the same as for the flexural shear determination discussed in Sec. 117. We take the shear flow at points P1. P2 as the redundants,
=
and express the shear flow as
J = 1, 2
(1377)
+ ii.
where Zj0 is the open section distribution and
(1378)
is due to The distribution, has the same form as We just have to replace C with C'S. We generate by integrating (i) around the centerline, and enforcing equilibrium at the junction points. For example, at point b (see Fig. 137),
+ = Note that = 0 at points P1, P2. e andf The redundant shear flows are evaluated by requiring no energy coupling between qU and qr which is equivalent to requiring qr to lead to no twist deformation, j. Noting (c), we can write
=
Finally, substituting for we obtain
j = 1,2
(1379)
aCr = B
1'
dS
(1380)
414
Once 4) and zir are known, the cross-sectional properties (1 , Ci., X2r, xar) can be evaluated. Also we can readily generalize the above approach for an
n-cell section.
1
39.
In this section, we establish the governing equations for geometrically nonlinear restrained torsion by applying Reissner's principle. This approach is a mixed formulation, i.e., one introduces expansions for both stresses and dis-
placements. The linear case was treated in Sec. 135. To extend the formulation
into the geometrically nonlinear realm is straightforward. One has only to introduce the appropriate nonlinear strain-displacement relations. Our starting point is the stationary requirement t
V*)d(vol.)
d(surface area)]
a, are independent variables, C and b are where e(u), prescribed. We take the displacement expansions according to (133) and use the strain-
V*(o.),
= U1 + C02X3
W3X2
/4)
cn1(x3
x3)
U3 =
+ w1(x2
= Yiz =
Y13
(1381
The in-plane strain measures (62, 63, Y23) are of 0(w2), which is negligible
noting the definition equations for the force parameters, the first term in (a)
expands to
d(vol.) =
1
1)]
W3 + + F2[u,2. i + F3[u53 S + t+ + M2{w2, 1 co1, 1(u52 i + x3w1, + M3[w3, i w1, i T2w1. 1)] M0f1 + MRf +
1
i+
MQW1W1, 1}dxj
f See Eqs. 1333 and corresponding footnote. We are working with Kirchhoff Stress and
Lagrangian strain here. See Sec. 103. Eq. 1028. Tile displacement expansions assume small-finite rotation, i.e., sin w and cos w 1. To be consistent, we must use (1028).
SEC. 139.
GOVERNING EQUATIONS
415
The terms involving the external forces have the same form as for the linear
case, but we list them again here for convenience (see (136)):
+ F1u1 + where the end forces (the barred quantities) are defined as previously, for
example,
(1383)
(5Jp1
etc.
force parameters and to expand V*. In the linear case, there are 8 force
M3, and M4, MR. Two additional force measures (Me, MQ) are present for the nonlinear case but they can be related to the previous force measures. We proceed as follows. We use the stress expansions employed for the linear case with = They are summarized below for convenience
measures. F1
(see Sec. 13--5):
a11
F1
M2
1-x3 -TX2
M3
13
+ &ij
+
MT
+
_.-
Il equations for
where 4,, f, q, h2 and h3 are functions of x2, x3. Introducing (a) in the definition
and MQ leads to
= =
$2
f11F1
+ fl2M2 + fl3M3 +
+
+
if if
if
4,
=
(13-84)
$3 =
/34,
416
andt
MQ
+
$J(x2h2k
+ +xlh3k)dA
+ (k =
2, 3)
(1385)
+ = Certain coefficients vanish if the cross section has an axis of symmetry4 One can readily verify that
fi1F1
MQ (1386)
when the section is doubly symmetric. For generality, we will retain all the
terms here.
The complementary energy density function has the same form as for the
linear case:
----'
2Ek.,A
'2
+ -
13)
+
1 + '
+ +
((Mw +
+ X2rF3)
We have shown that it is quite reasonable to neglect transverse shear deformation due to warping (C. X2r X3r = 0) for a thin-walled open section. Substituting Equations (1382)(1387) in Reissner's functional and re-
F1,1 + b1 = 0
j+
w1F3
w1
1M2} + b2
0
0
+ F3 + w1F2 wi,1M3} + b3 =
1
1
(1 +
+ (1 +
+
1
2J32w1,
= 0 + m2 = M3,1 + F2 + m3 =
M2, 1 F3
0 0
SEC. 139.
GOVERNING EQUATIONS
417
where
Relations
1+
1+
1 + Wj,
1
i x2uS3,
w3
i)
1FF2
+
+
F3
X3r
1FF2
1 MrTJ =
+
+
(02
1+
(1388)
= =
(02,1
+ (0l,j(US2,j + /32(01,1)
+ +
j
+ /33(01,1)
= 1.
[CrM?+ X3rF2 +
Boundarp
= I' +
Conditions (+ for x1 = L,
for x1 = 0)
+ (I +
1+
= MT
These equations simplify considerably when the cross section is symmetric and transverse shear deformation is neglected.1' We discuss the general solution of (1388) in Chapter 18. The following example treats one of the cases, a member subjected to an axial force and torsional moment.
t See Prob. 1313.
418
CHAP. 13
Example 137
We consider a prismatic member (see Fig. E137A) having a doubly symmetric cross section, fully restrained at one end and loaded by an axial force P and torsional moment M. We are interested here in evaluating the influence of axial force on the torsional behavior. The linear solution (with no axial force) was derived in Example 13i.
x2
Fig. E137A
(M1 +
i) =
Force-Displacement Relations
= GJw11
= ErI,ji
F1 =
i
Boundary conditions
.xi=O
xj = L
F1
M1 + J3tF1w1,1 = Al
Integrating the last two equations in (a) and noting the boundary conditions, lead to
F1 =
M1 + /31F1w1,
const
= corlst
=P =M
SEC. 139.
GOVERNING EQUATIONS
419
where
P11
7;:ij
2GJ
GJA
iP
I+
+ F)
This expression reduces to Equation (g) of Sec. 136 when P = 0. Once f is known, we can determine the rotation by integrating (d), which expands to
+ F +
when we substitute for M1 using (b). The general solution is,
M
f
M
f=
[GJ
C1
(i
+
+
C3
+ Mx {i +
{C1
sinh px + C2 cosh
(We drop the subscript on x1 for convenience.) Finally, specializing (g) for these particular boundary conditions result in
f=
wi =
1 + cosh
tanh
sinh
{sinh jtx + (1 cosh
These equations reduce to (13-57) when P = 0. A tensile force (P > 0) increases the torsional stiffness whereas a compressive force (P < 0) decreases the stiffness. Equation (h) shows that the limiting value of P is 1. We let F, represent the critical axial force and the corresponding axial stress
11
(;J
to be less than the yield stress, (J/11) must be small with respect to unity. As an illustration, consider the section shown in Fig. E137B. The various coefficients (see Example 134) are
In order for
J=
420
x3
X2
and
r,,
G
(t'\2(
REFERENCES
1.
2.
3.
4.
5. 6. 7.
T., and N. B. CHRISTENSEN: "Methods of Analysis of Torsion with Variable Twist," J. Aero. Sci., pp. 110124, April 1944. TIMOSHENKO, S. J:: "Theory of Bending, Torsion and Buckling of Thin-Walled Members of Open Cross Section," J. Franklin Inst., pp.559609, 1945. VON KARMAN, T.. and W. C. CHIEN: "Torsion with Variable Twist," J. Aero. Sci., Vol. 13, No. 10, pp. 503510, October 1946. BENSCOTER, S. U.: "Secondary Stresses in Thin-Walled Beams with Closed Cross Sections," NACATN 2529, Washington, D. C., 1951. BENSCOTER, S. U.: "A Theory of Torsion Bending for Multiceil Beams," J. Appi. Mech., Vol 21, No. 1, 1954.
VON
VLASOV, V. Z.: Thin Walled Elastic Brains, israel Program for Scientific Translations,
Office of Technical Services, IJ.S. Dept. of Commerce, Washington. D.C. 1961. HEILIG, R.: "Der Schuberverformungseinfiuss auf die Wlbkrafttorsion Von Stilben mit offenern Profil," Der Stahlbau, April 1961. 8. HEILIG, R,: "l3eitrag zur Theorie der Kastentrhger beliehiger Der Stahlbau, December 1961. 9. J. T.: Mechanics of Elastic Structures, McGraw-Hill. New York, 1967.
10.
Ii.
1968.
12,
MAISEL, B. I.: "Review of Literature Related to the Analysis and Design of ThinWalled Beams," Technical Report 440, Cement and Concrete Association, London,
July 1970.
PROBLEMS
13.
421
DABROWSKi, R.: "Gekrinmte dUnnwandige Trager," Springer-Verlag, Berlin, 1968. GALAMBOS, T. V.: Structural Members and Fiames, Prentice-Hall, 1968. BLEICH, F.: Buckling Strength of Metal Structures, McGraw-Hill, New York, 1952. BURGERMEISTER, G., and H. STEin': Srabilitar Theorie, Part 1. Akademie Verlag, Berlin, 1957.
18.
CHILVER, A. H.: Thin- Walled Structures, Chatto and Windus, London, 1967. REISSER, E.: "Note on Torsion with Variable Twist." .J. AppI. Mech., Vol. 23, No. 2,
pp. 315316, June 1956,
PROBLEMS
131.
=
where
F2
13
2
(733
13
x2
(in A)
(onS)
This result applies when the cross section is assumed to be rigid with respect to in-plane deformation. The coordinate of the shear center is defined by
X3
if
X3
where
132. 133.
is the St. Venant torsional warping function. Hint: See Prob. 1111
Verify (1340) and (1344). This problem reviews the subject of the chapter in two aspects.
X3 + X3]
=
(1339). Verify that
+ x2
x2]
MR when & =
422
When the cross section is thin-walled, (a) and (b) take the form
fquqr_ = o
is the perpendicular distance from the shear center to the tangent at the centerline, Equation (d) follows from (1129) and
where
Prob. 114. We determine qf from (1343). Finally, the force parameters for the thin-walled case are defined as
= MR =
Verify that
1.
dS
Jqrc&
dS
2. 3.
134.
I and compare vs. Mu. Also evaluate L and compare with the unrestrained value. 135. Refer to Examples 122 and 132. Discuss how you would modify the member force-displacement relations developed in Example 122 to account for restrained torsion. Consider 1, X3r = 0, and (a) warping restrained at both ends (b) warping restrained only at x L 136. Refer to Example 132. Determine the translations of the shear center. Consider the cross section fixed at x 0. Discuss how the solution has to be modified when the cross section at x = L is restrained against translation. 137. Starting with the force-deformation relations based on the mixed formulation (1349), derive the member force-displacement relations (see Example 122). Consider no warping at the end sections and take = + 1. Specialize for (a) symmetrical cross section (b) no shear deformation due to restrained torsion and flexurearbitrary cross section,
at x
13S.
material properties (F, G). Discuss how the displacement and mixed formulations haveto be modified to account for variable material properties. Note: The unrestrained torsion and flexural stress distributions are treated in
Prob. 1114 and 121.
139.
Determine the distribution of qr, and expressions for Cr, for the cross sections shown in parts a and b and part ed of the accompanying sketch (four different sets of data). and qr for the section shown. 1310. Determine 1311. Using the fiexural shear distributions listed in Prob. 131, show that
-
'12 =
PROBLEMS
423
Prob. 139
Ii
T
F 0.75k
(b)
I I
2k
(c)
See part c.
(d)
/z
'iH
Prob. 13-10
t
0
t
I
a
0
1s2
I
424
(x2 V q52r +
22
113
1312. Specialize Equations (1384) and (1385) for the case where the cross section is symmetrical with respect to the X2 axis. Utilize
x3)H0(x2, x3)dA =
where He is an even function and H,, an odd function of x3. Evaluate the coefficients for the channel section of Example 135. Finally, specialize the equations for a doubly symmetric section. 1313. Specialize (1388) for a doubly symmetrical cross Section. Then specialize further for negligible transverse shear deformation due to flexure and warping. The symmetry reductions are
X2
=
!72
=0
X2r
X3r = 0
= Consider the two following problems involving doubly symmetric cross section. (a) Establish "linearized" incremental equations by operating on (1388)
1314.
i/A23=O 'li 0
Specialize for a doubly symmetric cross section (see Prob. 1312). Consider the case where the cross section is doubly symmetric and the initial state is pure compression (F1 P). Determine the critical
co1 =
==
at x =
at x
0,
(restrained warping)
2.
=0
0, L
(unrestrained warping)
Neutral equilibrium (buckling) is defined as the existence of a nontrivial solution of the linearized incremental equations for the same external load. One sets
F1 =
U2
P
0
U3 = W1 = (02 = (03 = f
and determines the value of P for which a nontrivial solution which satisfies the boundary conditions is possible. Employ the notation
introduced in Example 137.
Determine the form of V, the strain energy density function (strain energy per unit length along the centroidal axis), expressed in terms of displace1315.
ments. Assume no initial strain but allow for geometric nonlinearity. Note that V = V* when there is no initial strain.
14
Planar Deformation of a
Planar Member
14i.
2.
3.
The plane containing the centroidal axis also contains one of the
principal inertia axes for the cross section. The shear center axis coincides with or is parallel to the axis. However, the present discussion will be limited to the case where the shear center axis lies in the plane containing the centroidal axis.
We consider the centroidal axis to he defined with respect to a global reference frame having directions X1 and K2. '[his is shown in Fig. 14i. The orthogonal unit vectors defining the orientation of the local frame (Y1, Y2) at a point are x 12 = where points in the positive tangent direction and denoted by 13. Item 2 requires Y2 to be a principal inertia axis for the cross section.
x2
Yl n
r2
tl
B
S
A
n
i2
x1
ii
Fig. 14i.
426
CHAP. 14
By definition, t
dx1
dx2
(14-1)
13, it
follows that
dx5
(142)
(14-3)
where
1
dt1
d2x1 dx2
d2x2 dx1
According to this definition, R is negative when d11/dS points in the negative t2 direction, e.g., for segment AB in Fig. 141. One could take t2 = ii, the unit normal vector defined by
1
d11
(14-4)
ciS
x 12 = 13 but this choice is inconvenient when there is a reversal in curvature. Also, this definition degenerates at an inflection point, i.e., when dt/dS = O. If the sense of the curvature is constant, one can always orient the X1-X2 frame so that coincides with , to avoid working with a negative R. To complete the geometrical treatment, we consider the general parametric representation for the curve defining the centroidal axis,
x1 = x1(y) x2 = x2(y)
where y is a parameter. The differential arc length is related to dy by
dS
2 2
(j45
1/2
d. -
+ (p)]
dv
dy
(146)
SEC. 142.
FORCE-EQUILIBRIUM EQUATIONS
427
t1 = (
I 7dx1
1j
t2 = (
1
if
1
+ dy
+
dx1
dx2
dx2..
---
dy
dy
(14_
if._ (It1 = -( t2 dy
(
k\
dy
d2x2 dx1
dy2 dy
A planar member subjected to in-plane forces plane for our notation) will experience oniy in-plane deformation. In what follows, we develop the governing equations for planar deformation of an arbitrary planar member. This formulation is restricted to the linear geometric case. The two basic
solution procedures, namely, the displacement and force methods, are described and applied to a circular member.
We also present a simplified formulation (Marguerre's equations) which is valid for a shallow member. Finally, we include a discussion of numerical integration techniques, since one must resort to numerical integration when
the cross section is not constant.
142.
FORCE-EQUILIBRIUM EQUATIONS
The notation associated with a positive normal cross section, i.e., a cross
section whose outward normal points in the + S direction, is shown in Fig. 142.
We use the same notation as for the prismatic case, except that now the vector
- I'3
dA
012
Centroidal axis
Fig. 142. Force and moment components acting on a positive cross section.
428
CHAP. 14
components are with respect to the local frame (Y1, Y2, Y3) rather than the
basic frame (X1, X2, X3). The cross-sectional properties are defined by
A
13
if dy2 dy2
if dii
'2 = .iJ(Y3)2
= JJ(y2)2 dii
(148)
Since
Y2, Y3 pass through the centroid and are principal directions, it follows
that
dA = flY3 dA =
loading,
F3
M1
SSY2Y3
dii =
(149)
(1410)
in this case, we work with reduced expressions for F and M (see Fig. 143) and drop the subscript on M3:
M+ =
M3t3
+ F212' = Mt3
(14-11)
)
= t1 x t2
x1
To establish the force-equilibrium equations, we consider the differential volume element shown in Fig. 144. We define b and as the statically equivalent external force and moment vectors per unit arc length acting at the centroid.
For equilibrium, the resultant force and moment vectors must vanish. These conditions lead to the following vector differential equilibrium equations:
dS
dM
,_.
in + r1 x F+ =
(1412)
SEC. 143.
429
We
expand b and
b= + = mt3
b212
(1413)
Introducing the component expansions in (1412), and using the differentiation formulas for the unit vectors (143), lead to the following scalar differential equilibrium equations:
dF1
F2
+ b1 = 0
(14-14)
dM
+m
that the force-equilibrium equations are coupled due to the curvature. The moment equilibrium equation has the same form as for the prismatic case.
dS
r(S)
The positive sense of the end forces is shown in Fig. 145. We work with components referred to the local frame at each end. The end forces are related to the stress resultants and stress couples by
= = Mj52 =
MA= MISA
(14-15)
j=1,2
143.
We establish the force-displacement relations by applying the principal of virtual forces to a differential element. The procedure is the same as for the
430
CHAP. 14
prismatic case described in Sec. 123, except that now we work with displacement components referred to the local frame at each point. We define and as
= =
For planar deformation, only u1, u2 and 0J3 are finite, and the terms involving
u3, co1, and w2 can be deleted:
u1t1 + U2T2
C03t3
Wt3
(1417)
F41
x2
x1
SEC. 143.
431
We define
deformation, = (F1, F2, M). One determines by taking expansions for the stresses in terms of F1, F2, M, substituting in the complementary energy density, and integrating with respect to the cross-sectional coordinates Y2, y3. We will discuss the determination of later.
Specializing the three-dimensional principle of virtual forces for the onedimensional elastic case, and writing
cF1
e1
AF1
0F2
AF2
k
cM
AM
(1418)
AF1 + e2 AF2 +
k
AM
+ e2 AF2 +
AM)dS
AP1
(1419)
where is a displacement measure and is the force measure corresponding to d1. The virtual-force system (AF1, AF2, AM, AP1) must be statically permis-
We apply (1419) to the differential element shown in Fig. 147. The virtual
AF =
(a)
Evaluating
AP1,
= = {AF1
+AM
+ AF2
+
(b)
dS
432
CHAP. 14
and then substituting in (14-49) results in the following relations between the
du2
u1
(14-20)
eV*
k
dw dS
ar2
dy =
di2
dt-\
dv
Noting that
dy
(112
7k-ti
dv
By definition,
dS2 dv2 dv
(142 1)
if
.-
In general, V* = V" (11, We select suitable expansions for the stress components in terms of F1, F2, M, expand V*, and integrate over the cross section. The only restriction on the stress expansions is that they satisfy the definition equations for the stress resultants and couples identically:
dA
SEC. 143.
433
(1422)
where I 13. A logical choice for (when the cross section is thin-walled) is the distribution predicted by the engineering theory of flexural shear stress
distribution described in Sec. 117: a11 = 1q(F2)
q=
F2t/i
(1423)
where t denotes the local thickness, and q is the flexural shear flow due to F2. Both expansions satisfy (a).
x2
r +1)212 +Y33
r2
r1ty +dy)
Y2
it
In what follows, we consider the material to be linearly elastic. The complementary energy density is given by
11*.....
a12
where c? is the initial extensional strain. Substituting (a) in (1421) and taking
434
CHAP. 14
for V*:
= e?Fi + kM +
where
+ =
+
dA
2GA2*
(14-24)
if
55
(i
I
e,
\\
R}
=
are
A2.
1 and If the section is symmetrical with respect to the 1'3 axis, 1* The deformation-force relations correspoiiding to this choice for
ei
F2
F1
M
du2
U1
dr,1
u2
=
F1
-w
dw
(14-25)
Note that the axial force and moment are coupled, due to the curvature.
Inverting (1425) leads to expressions for the forces in terms of the deformations:
F1 =
M
LA
e1)
R(l
EI*
R(1
)(el
(14-26)
F2
We observe that
where p is the radius of gyration and d is the depth of the cross section, For
example,
1
d2
AR2 = i2R2 for a rectangular cross section. Then, is of the order of (d/R2) and can be
neglected when (dIR)2 1. A curved member is said to be thin when O(d/R)
1.
1
We set =
SEC. 144.
435
neglecting y2/R with respect to unity in the expression for the differential arc length, i.e., by taking
by
dS
14 27
for V* developed for a prismatic member. The approximate form of (1425) for a thin member is
F1
dii1
Li2
(1428)
ik
To complete the treatment of the linear elastic case, we list the expanded forms of the principle of virtual forces for thick and thin members. Note that these expressions are based on a linear variation in normal stress over the cross
section.
Thick Member
Cit0
F1
M'\ AF1
+
F2
/XF2
(1429)
+ (ko +
Thin
AM} dS
dS =
(1430)
d1 AP1
144.
In the variational procedure for establishing one-dimensional force-displacement relations, it is not necessary to analyze the deformation, i.e., to determine
the strains at a point. One has only to introduce suitable expansions for the stress components in terms of the one-dimensional force parameters. Ndw, we can also establish force-displacement relations by starting with expansions for the displacement components in terms of one-dimensional displacement parameters and determining the corresponding strain distribution. We express the
436
CHAP. 14
stresses in
terms of the displacement parameters using the stress-strain relations, and then substitute the stress expansions in the definition equations for F1, F2, and M. The effect of transverse shear deformation is usually neglected in this
QQ2 =
dy2
dy2 t2
a
(1431)
a2 =
We use a prime superscript to denote qua Iltities associated with the deformed position of the member, which is shown in Fig. 1410; for example:
?'=
= position vector to point P(y) in the deformed position (point P'). tangent vector to the deformed centroidal axis. = position vector to Q(y, Y2, y3) in the deformed position (point Q').
x2
Q2
Q1
Pj(y +dy)
P(y)
axis
x1
SEC. 144.
437
P'P'1 =
0))
+
C))J
=
dy
dy2
(1432)
or2
&Y2
/
\.
c,u2
(j =
1,
'12 3
Sin Y12
(1-3)
Now, we restrict this discussion to small strain, Substituting for the deformed
Istj,
au2
c'y
+
2
2(a2)"
,V
1 ('U2
Y12
tj
-'
"
a2
t2
cc2
The nonlinear terms arc associated with the rotation of the tangent vector. Neglecting these terms corresponds to neglecting the difference between the
deformed and undeformed geometry, i.e.. to assuming linear geometry. The next step involves introducing an expansion for in terms of y2. We express 2 as a linear function of
wv211
(1435)
(1436)
is the displacement vector for a point on the centroidal axis. Equation (1435) implies that a normal cross section remains a plane after deformation. One can interpret co as the rotation of the cross section in the direction from toward t2. This notation is illustrated in Fig. 141 1. In what follows, we consider only linear geometry. Substituting for 2, taking y = S, and evaluating the derivatives lead to the following strain expansions:
438
CHAP. 14
y2k)
e2
= du,
u2
= I61IY20
0)
=
doi
(1437)
The vanishing of c2 is due to our choice for 2. One could include an addiThis would give tional linear term, = $ and, additional terms in the
x2
Q2
x1
(u1 Wy2)tl
Centroidal axis
UI tl
SEC. 144.
439
expressions for
section remains plane does not lead to a linear variation in extensional strain over the depth when the member is curved. We introduce the assumption of negligible transverse deformation by setting e2 = 0. The resulting expressions for (0 and k in terms of u1 and u2 are
e2
du2
dS
u1
R
dIui
dS2
(1438)
dS
When transverse shear deformation is neglected, one must determine F2 using the moment-equilibrium equation. The next step involves expressing F1, F,, and M in terms of the one-dimensional deformation parameters e1e2 and k. In what follows, we consider the material to be linearly elastic and take the stress-strain relations for c12
as:
E(c1
= Gy12
---(e1
F 1 y2/R
y,k)
Fe1
(1439)
F2
= Ge2 if =
Fe1
d
(14-40)
+ Ekjj
The various integrals can be expressed in terms of only one integral by using the identity
1
1 y2/R
1-F
1
y2/R
axis:
$5Y2 dA
0
11
= (733
We generally neglect
for a
440
CHAP. 14
ri
c/A
I'
dA
L
JJ
1
(1441)
- y2/R
+
F2
F1
= k +
lvi
where
=
= A(1
442
e? =
if
(i
c/A
k =
if
dA
The expressions for e1 are identical with the result (see (1425)) obtained with
the variational approach. However, the result for k differs in the coefficient for M. This difference (1' or F') is due to the nonlinear expansion used for
Example 141
We determine I' for the rectangular cross section shown in Fig. F.141.
I' =
11
1
y2/R
=h
Ja;2 1 y2/R
=R2bd+R3bln
To obtain a more tractable form, we expand the log terms, using
(1+x'\
SEC. 144.
441
In
d3
ii
d
3(d\2
+ +
3(d\4
+
and
I' =
{
I
+ ..
Fig. E141
H
Y3
The relations listed above involve exactintegrals. Now, when the member
is thick, we neglect (y2/R)2 with respect to unity. This assumption is introduced by taking
+ and I':
+ ...
Co
.--..e2
Y2
- yJR
JJ
dA
=i{i
442
CHAP. 14
To be consistent, we must also neglect 1'/AR2 with respect to unity in the expression for A'2 and I". When the member is thin, we neglect y2/R with respect to unity.
1y2/R
y2k
(1444)
at2
It is of interest to establish the one-dimensional form of the principle of virtual displacements corresponding to the linear displacement expansion used in this development. The general three-dimensional form for an orthogonal coordinate
system is (see Sec. 106):
SJJ(aii
where
+ a12 y12
)d(vol.) =
represents an external force quantity and d1 is the displacement quantity We consider only and Viz to be finitc, and express the corresponding to differential volume in terms of the cross-sectional coordinates Y2' Y3 and arc length along the ccntroidal axes (see Fig. 149):
d(vol.) =
dS2
dy2 dv3
(i
dS
d7 dy3
+ a12
(i
dA] dS =
(1445)
We take (1445) as the form of the principle of virtual displacements for planar
deformation.
The strains corresponding to a linear expansion for displacements and linear geometry are defined by (1437), which are listed below for convenience:
ci
e2
Y12
y2/R
U2
du1
du2
=
k
u1
U)
do
dS
Substituting for e1, Y12 and using the definition equations for F1, F2, and M,
SEC. 144.
443
we
obtain
Ad1 (1446) + F2 + M k]dS = Js[Fi This result depends only on the strain expansions, i.e., (c). One can apply it for the geometrically nonlinear case, provided that (cS) are taken as defining
the strain distribution over the cross section. We use the principle of virtual displacements to establish consistent forceequilibrium equations. One starts with one-dimensional deformation-displacement relations, substitutes in (1446), and integrates the left-hand side by parts. Equating coefficients of the displacement parameters leads to a set of force equilibrium equations and boundary conditions that are consistent with the geometrical assumptions introduced in establishing the deformation-displacement relations. The following example illustrates this application.
Example
142
equal to zero. This leads to an expression for the rotation. w, in terms of the translation
components,
(In2
hEFt e1 + M k]dS
1< =
d (du2
u1
= Au1 id
Au2
1
Aui
Au2
AU2
(5k
d2
and integrating by parts, the left- and right-hand sides of (b) expand to
j 54
[F1
+ M (5k]dS
/ F1 /
I
Rj
M\
Au1
dM
dS
Au2
+M
d uS
An2
F1 +
I
+j
1Aui
Rj r
Au1 -- Au2
dM
dS
1
+M
dS
dF1
dM1
+ Au2
[
[
F1
dS
444
CHAP. 14
and
Ad,
(b1 +
(b2
dS
+ (PB! +
+ (p41 +
+ (r42
014) AUA2 + M4 A
The consistent equilibrium equations and boundary conditions for negligible transverse shear deformation follow by equating corresponding coefficients of the displacement variations in (e) and (f):
S4<S<S3
+
dF1
F1
1dM
d2M
+ b1 +
01
+-
drn
=0
ssn
u1
F1
u2
p42
do2
M=
MA
S=
UI
U2
F1 =
=
M
F21
F32
in
du2
One can obtain (g) by solving the last equation in (1414) for F2 and substituting in the first two equations. Suppose we neglect u1/R in the expression for w:
do2
CD
d2u2
t See Ref. 5.
SEC. 145.
CARTESIAN FORMULATION
445
The other equilibrium equation and the boundary conditions are not changed. Using (h) instead of (a) eliminates the shear term, F2/R, in the tangential force-equilibrium
equation.
145.
CARTESIAN FORMULATION
We consider the case where the equation defining the centroidal axis has the form x2 = f(x1). The geometrical relations for this parametric representation are obtained by taking y x1 in (147). They are summarized belowt for convenience and the notation is shown in Fig. 1412:
dS =
dx1
= i[
I
r [
I
7df\21112
ylx1jj. f'df'\ + I)
-
cos0
1
j
t2
[ / df \
d2f
+ '2
(1447)
= t1 X t2 = 13
ci:
In the previous formulation, we worked with displacement components and external force components referred to the local frame. An alternate approach, originally suggested by involves working with components referred to the basic frame rather than the local frame. The resulting expressions differ, and it is therefore of interest to describe this approach in detail. We start with the determination of the force-equilibrium equations.
Consider the differential clement shown in Fig. 1413. The vector, equilibrium
equations are
dF+
dx1
See Prob. 141. See Ref. 6.
(1448)
ax1
x
dx1
446
CHAP. 14
x2
Y2
YI
X2
dx1
x1
'I
dx1
pN2 12
1
F1t1
lj
Fig. 1413. Differential element for equilibrium analysis.
SEC. 145.
CARTESIAN FORMULATION
447
h are the external applied force and moment vectors per unit projected length, i.e., per unit x1. They are related to b and (see Fig. 144) by
where fl,
(1449)
(1450)
sin 0) =
P2
dx1
dx1
(F1
sin0 + F2 cos 0)
(14-51)
1(dM
'\
F2
N1 sm0 + N2cosO
AF1 + e2 AF2 + k
dx1 =
d1
(a)
where V* V* (F1, F2 M) is the complementary energy per unit arc length. Consider the differential element shown in Fig. 1414. The virtual-force system is statically permissible, i.e., it satisfies the force-equilibrium equations identically:
dx1
dx1
= =
+ J1 x
Expanding
d1
+ 1)212
(013 = (0t3
(1452)
448
CHAP. 14
we
obtain
dx1; Finally, substituting for N1, N2, in terms of F1, F2 and equating coefficients of the force increments result in
dx1 dx1
(d)
oV
cos2 0- + sinOcosO.
do1
do2
dx1
e2
a v*
sin 0 COS
do1
U
--
+cosOw dx1
do2
(1453)
k=
= cos 0
dw
dx1
The member is said to be shallow when 02 << 1. One introduces this assumption by setting 4f cos
U
1
Sm 0
tan 0 =
(1454)
1)2
2
V1
Marguerre's equations are obtained by assuming the member is shallow and, in addition, neglecting the contribution of F2 in the expression for N1.
SEC. 146.
449
N,
N2
F1
F2 +
dF, + Pi = dx,
dF2
dx,
F2
d 7
dx, dM
dx1
(F1
df\
dx,j
J
+ P2 =
in (1455)
e, =
e2
dv'
d.x,
dv2
0F2
OM
= - co
dx,
dw
dx,
One step remains, namely, to establish the boundary conditions. The general conditions are v1 or N, v2 or N2 prescribed at each end (1456) M or w
We obtain the appropriate boundary conditions for the various cases considered above by substituting for N,, N2 and ox For example, the boundary conditions for the Marguerre formulation are
or
w
F2 +
M
(14-57)
or
146.
The
ferential equations which, for the planar case, consist of three force-equilibrium
equations and three force-displacement equations. If the applied loads are independent of the displacements, we can first solve the force equilibrium equations and then integrate the force-displacement relations. This method is quite straightforward for the prismatic case since stretching and flexure are uncoupled. However, it is usually quite difficult to apply when the member is
450
CHAP. 14
curved (except when it is circular) or the cross section varies. In what follows,
we illustrate the application of the displacement method to a circular member having a constant cross section, starting with
1.
2.
the exact equations (based on stress expansions) for a thick member Marguerre's equations for a thin member
The results obtained for this simple geometry provide us with some insight as to the relative importance of transverse shear deformation and stretching deformation versus bending deformation. When the centroidal axis is a circular segment, R = const, and the equations simplify somewhat. It is convenient to take the polar angle 8 as the independent
variable in this case. We list the governing equations below for convenience and summarize the notation in Fig. 1415:
dF1
dM
/
=
R2b2
RF1
(1458)
1dM
= ej +
Eq. (1425)
e2
F1
1 f'du1
+ ui)
F1
0)
(1459)
k=
Idw
RF1 = M
R2
(b1
+ M
C1 + R2[b2
M=
C1
+ C2cosO + C3sinO +
(1460)
SEC. 146.
451
where
denotes the particular solution due to the external distributed loading and C2, C3 are constants. Once M is known, we find F1 using (a) and F2 from the moment equilibrium equation. The resulting expressions are
F1 =
F2 =
j(b1
in
;)dU
(1461)
dS
RdO
F1
Re? +
RF2
+ RF1)
du2
+ u1
+ Rco
= Rk +
To determine u1 and u2,
[M +
(RF1)]
= u2 + Re? +
(M + RF1)
(1462)
452
CHAP. 14
and
d2u2
= =
1/1
Re?
+ RF1)
(1463)
R2k
+ a1R2
Re?
I
a1 =
to neglect with respect to 1 but we will retain it in order to keep track of the influence of extensional deformation. We solve (1463) for u2, determine u1 from (1462), and w from the second equation in (a),
F2
W
1
(du,
'\
(14-64)
This leads to three additional integration constants. The six constants are determined by enforcing the three boundary conditions at each end. Various
loading conditions are treated in the following examples.
Example
to
143
Consider a member (Fig. El43) fixed at the negative end (A) and subjected only
at the right end (B). The boundary conditions for this case are
F1=Fa1;
F2=M=O u1 =u2=w=O
atO=011
atO=O
Specializing the force solution for no external distributed loading and enforcing the boundary conditions at B, we obtain
F1 = F2 =
M
F81
cos(08 0)
8)
FRI Sifl(08
RF51(l cos(08
0))
To simplify the analysis, we suppose there is no initial deformation. Using (b). the form
F81
takes
R'
'I'
[a1
02
COS(08
0)1
where
EI*
(d\'
SEC. 146.
453
Note that
c4
cos
0 + c5
sin o +
[ai +
0)]
=
+
sin 0 C5 cos 0 + C6
{o +
[o
0)
+ sin(66 0)]}
{O
sm(OB O)}
Finally, the constants are found by enforcing the displacement boundary conditions at 0 = 0:
C4 =
C5 = C6 =
To determine the relative importance of stretching and shear deformation versus bending deformation, we evaluate the displacements at 0 and write the resulting expressions
Fig. E143
A
F R3
(05
sin
FBI
454
CHAP. 14
WE =
UBI
sin
+ b1 Oe)
b2
+ b3 &)
4Sin
sin
Slfl 63
b2
(I)
4 03
63 +
sin 03 Co
T
4sin2 63
I + COS 63
The coefficients (b1,..., b4) are of order unity or less when is not small with respect to unity, i.e., when the segment is not shallow. Also, e and , are of order (d/R)2. It follows that the displacements due to stretching and shear deformation are of order (dIR)2 times the displacement due to bending deformation for a nonshallow member. To investigate the shallow case, we replace the trigometric terms in (i) by their Taylor series expansions,
sin 0 =
02
02
cos 6 = I sin0cos6
0(1
and neglect
1*
U3j
[1
EI*
Now,
1*
El"
=
(d'12
SEC. 146.
455
For example,
1 (d\2
AS2
EI*
E (d\2
lOG
(d'\2
0,26
for a rectangular section and v = 0.3. Since (cl/S)2 << I for a member, we can neglect the transverse shcar terms in UBL, UB2 and the stretching term in co8. However, we must retain the stretching term in u51 since it is of the same order as the bending term. The appropriate expression for is PatS3
In sum, we have shown that the percentage of error due to neglecting stretching and transverse shear deformation is of the order of (d/R)2 for a nonshallow circular member. If the member is shallow < I 5), we catnwt neglect stretching deformation. Actually, the stretching term dominates when the member is quite shallow. The error due to neglecting transverse shear deformation for the shallow case is still only of the order of
(d/R)2.
Example
144
F2 = F52
cos(05
0) 0)
R sin(02
We suppose the member is not shallow and neglect stretching and shear deformation. The force-displacement relations reduce to (we set A = = in (1459))
du1
dO du2
= Re?
u1
= Rw = Rk + EJ*
RM
dw
dO
+ u2 =
du1
R2k
u2
I
Re?
R2
+ Re?
(du2
We determine u2, then u1, and finally oi. Note that (c) corresponds to (1462), (1463) and (1464) with A = A2 = cc. The final expressions (for no initial deformation or support
456
CHAP. 14
movement) are
=
02
{(O COS(OB 0) sin
0 cos
0
0 Sjfl(OB 0) + COS(OB
=
0) =
F52R
{
cos
0)}
COS(OB 0) cos
Example
145
We analyze the shallow parabolic member shown in Fig. E145 using Marguerre's equations. We consider the member to be thin and neglect transverse shear deformation. Taking f = and = rn = 0, the governing equations (see (1455) and (1457)) reduce to
dx1
d2M
dx1
al'1 P2 = 0
F2 F1
dM
=
Cl)
e?
AL
= - + ax1
dx1
dv1
dx1
dv2
dx1
k=
v1,
d2v2
v2, w prescribed at x1
N2 =
dx1
+ ax1F1 = 0 at x1 = I.
M =0
Integrating (a) and using the boundary conditions at x1 = L, we obtain
M=
F2 =
We suppose e? = k = relation,
0
xi)2
x1)
x?)
p2(L
ax1N51
=M =
x1)2
CINBI(L2
4)
SEC. 146.
457
Fig. E144
l.A const
Fig. E145
,.- F1
a=
L2
Al
(h/L)2<(1
t
P2 =
COflSt
B
NB!
458
CHAP. 14
dv2/dxj =
at x1 =
Ely2 =
relation,
dv1
F1
dv2
NB1X1
AE
6 ) 1) [VT)
a
2h,/L2
Now,
Then
a2L4(A'\
6
2(h'\2
and
we see that this term dominatcs when h is larger with respect to the cross-sectional
depth.
147.
Our starting point is the principle of virtual forces restricted to planar deformation, Js(ei AF1 + e2 AF2 + k AM)dS AR1 = d1 (1465)
represents a support movement, and AR1 is the corresponding reaction increment. The relations
where the virtual-force system is statically permissible,
between the deformation measures (e1, e2, k) and the internal forces (F1, F2, M)
depend on the material properties and on whether one employs stress or displacement expansions. This discussion is limited to a linearly elastic material but one should note that (14--65) is valid for arbitrary material. For convenience, we list the force-deformation relations below. The notation for
internal force quantities is shown in Fig. 143.
F1
e2 =
F2
(1466)
k=
F1
AER.
+ -= El
SEC. 147.
459
k, A2, and] are defined by(1424) for the stress-expansion approach and (1442) for the displacement-expansion approach. Thin Linearly Elastic Member
e1 =
e?
F1
(14-67)
where A2, a?, k are the same as for a prismatic member. When the member is nor shallow, it is reasonable to neglect stretching and transverse shear deformation. As shown in ExampLe 143, this approximation
introduces a percentage error of O(cl/R)2. Formally, one sets A = A2 = If the member is shallow, we can still neglect transverse shear deformation but we must include stretching deformation. The basic steps involved in applying the force method to a curved member arc the same as for the prismatic case. However, the algebra is usually more complicated, due to the geometry. We will discuss first the determination of the displacement at a point. To determine the displacement at Q in the direction defined by tQ, we apply an external virtual force, APQ7Q, generate a statically determinate system of internal forces and reactions corresponding to = FJ,QAPQ AM = M AP0
ARk =
and substitute in (1465):
dQ = SS(eIFI,Q + e2F2Q + kM,Q)dS
(1469)
Rk.
Q
(j =
1,2)
(14-68)
APQ
This expression is valid for an arbitrary material. We set e2 = 0 if transverse a? if stretching deformation is
Example 146
We consider the thin linearly elastic circular segment shown in Fig. E146A. We
suppose the member is not shallow and neglect stretching and franslerse shear deformation. The reactions are the end forces at A for this example, and (1469) expands to
+ (k0 +
(a)
460
CHAP. 14
Fig. E146A
U2L
Expressions
Displacements at B
= AF51. The internal virtual-force system corresponds 0 as tlie independent variable =
rather than 0. The force-influence coefficients (F10, F2,Q, M.Q) follow directly from Fig. E146B:
F10 =
=
R(1
COSq
F20 =
=
cos
(b)
{e? cos
+ R (ko +
sin
(1 cos
cos
Jo3
cos
Once the loading is specified, we can evaluate the integral. Terms involving the support displacements define the rigid body displacement at B. leads to expression for u82 and Taking = SF82, We list them below for future reference:
"
u82 = R
Jo
COB =
(.
Elj
(ko +
(d)
W.4 + R
f3
SEC. 147.
461
Fig. E146B
ill
Ffi2 ,t482
RI
eQS
Fig. El 4SC
A.
_______________
462
CHAP. 14
= =
The expressions for the displacements at B due to an external loading are obtained by specializing (c) and (d) for no initial deformation or support movement and noting that
M=O
M=
RPc1[1
iic)] +
sin(1i
ilc) +
ivIc
Pc1R3(
=
Sifl 0c sin
+ Sill 1k +
sin
.
1k +
sin
Sill 0c cos
PC2R(
MCR2
cos
Or
sin
+
=
(Oc + Sin
OR + cos
Sin
1 1
PciR3I
+ --h-
Oc sin
sm Oc sin
PC2R3 /1
cos 1k
cos
sin 9c cos
+ -- (ens 1k
.
Sin Oc) +
R2Pc2
cos Oc) +
If we take point C to coincide with B, = 0 and = OR. The resulting equations relate the displacement at B due to forces applied at B in the directions of the local frame at B and can be interpreted as member force-deformation relations. It is convenient to express these relations in matrix form:
= R2[l cos
OR]
2 sin
OR
sin OR]
F81
Symmetrical
R[l
cos 98]
F82
We call
We describe next the application of the principle of virtual forces in the analysis of a statically indeterminate planar member. Let the member be indeterminate to the rth degree and let Z1, . , Z, represent the force redundants. Using the equilibrium equations, we express the internal forces and reactions in
.
SEC. 147.
463
terms
F1 = F10 +
k-=
1
F2 =
F2,0 +
(1470)
M = M0 +
k= 1
R1 = R10 +
k1
R1.kZk
(which is statically Substituting the virtual force system corresponding to permissible and in (1465) and letting j range from 1 to r lead to the compatibility equations relating the actual deformations:
+ kM,1)dS
(a)
1=
form
l,...,r
(j =
When the material is linearly elastic, the compatibility equations take the
k1
where
...,r)
+ Fl.kM,f)
(1471)
=fkj =
f
+
+
+
=
+
+ Fl,OM.J) +
,42,and 1/AR
Wesetl = I,A2 =
Oforathinmember.
Note that fik is the displacement of the primary structure in the direction of is the actual displacement of the point of Z1 due to a unit value of Zk. Also, minUs the displacement of the primary structure in the direcapplication of tion of Z, due to support movement, initial deformation, and the prescribed
external forces.
Example
147
Consider the symmetrical closed ring shown in Fig. E147. From symmetry,
at6r=0
F2 =
0
(a)
464 We take
CHAP. 14
the moment at 0 = 0 as the force redundant. To simplify the algebra, we suppose the member is thin and neglect stretching and shear deformation. The compatibility equations reduces to f11Z1 =
fii =
&=
due to a unit value of Z1 and is the relative rotation is the relative Note that rotation (X) due to the applied load. Equation (h) states that the net relative rotation
must vanish.
E147
fm/F
F1
1'
Now,
R(t cos 0)
dS
cos 8)dO
/PR\
JM21 cIS
PR1
TI\1
2\
)
SEC. 147.
465
Because of symmetry, we need to integrate over only a quarter of the ring. Finally, the total moment is
F2 =
When the equation defining the centroidal axis is expressed in the form x2 = f(x1), it is more convenient to work with force and displacement tities referred to the basic frame rather than to the local frame, i.e., to use the
cartesian formulation developed in Sec. (145). The cartesian notation is summarized in Fig. 1416.
xl
N2,
F2
--'-
i2
it
The geometrical quantities and relations between the internal force components are
tan 6 =
dS =
dx1
cos 6
F1 = F2 =
N1
(1472)
CHAP. 14
first find N1, N2 and then determine F1, F2. To obtain the equations for the Cartesian case, we just have to replace dS by dx1/cos U in the general expressions ((1469) and (1471)). In what follows, we suppose the member is thin and linearly elastic. When the member is not shallow, we can neglect the stretching and transverse shear deformation terms. The equations for this case reduce to:
We
Displacement at Point Q
dQ
+ (ke +
(14-73)
C'oinpatibility Equations
= j =
We can
\dx1
(1474)
5[e?Fi,i
+ (ko +
and N2
since
in terms of N1
Plo =
Then,
cos
0 N10j + sin 0
+ ,f'N2, (
(1475)
j'
tan 0 =
dx1
(1476)
F1
N1 +f'N2
f'N1 + N2
F2
We cannot neglect the stretching deformation term in this case. However, it is reasonable to take F1 N1. We also introduced this assumption in the development of Marguerre's equations. The equations for the shallow case with negligible transverse shear deformation and F1 N1 have the forms listed
below:
Displacement at Point Q
dQ
=J
[(ei +
N1,
+ (ko +
M,
dx1
R1,
(1477)
SEC. 147.
467
Compatibility Equation
>j1k4 =
fjk
=
=
+
Jx,
dx1
(1478)
+ (k0 +
Example 148
Consider the two-hinged arch shown in Fig. E148A. We work with reaction components referred to the basic frame and take the horizontal reaction at B as the force
redundant.
x2,
Fig. E148A
Primary
We must carry out two force analyses on the primary structure (Fig. E14SB), one for the external forces (condition Z1 = 0) and the other for Z1 1. The results are displayed in Figs. E148C and D, respectively.
Fig. E148B
V2
= Z1
zI
R242
468
CHAP. 14
Fig. E148C
N2
KM
0
Fig. E148D
(+)
H
'VI,'
ii
(+)
N2,,
M,1
Compatibility Equation
We suppose the member is not shallow. The compatibility equations for Z1 follow
from (1474):
I J11I El
dx t cosO
=
Jo L
+ f'N2,1) +
(ko
SEC. 147.
469
Using the results listed above, the various terms in (a) expand to
0E1 = .ft
=
cost?
dx1
s:
+ f'N2
+ k
JLIJLI(h)(h)
+ + +
JL[(
k0(f_
LIcosO
!L
(+P(xi
- a))dxi
(c)
Finally, the total forces are obtained by superposition of the two loadings:
M = M0 + Z1M1 R. = + Z1R,
R4 =
(d)
1,
2, 3
To evaluate the vertical displacement at point Q, we apply a unit vertical load at Q on the prinwry structure and determine the required internal forces and reactions plotted in
Fig. E148E.
Fig.
F
XQ
1/L
(i_IL)
FQ=+l
470
CHAP. 14
VA2
(13B2
VA2)
+
('XQ
x1Ik+1------x1
I'
M\
dx1
CL
I
\ 1L
dx1
A numerical procedure for evaluating these integrals is described in the next section.
Example
149
The symmetrical nonshallow two-hinged parabolic arch shown in Fig. E149A is subjected to a uniform load per unit horizontal length, that is, per unit x1. The equation for the centroidal axis is
4h(
where h is the elevation at mid-span
right end as the force redundant and consider only bending deformation. Figures E149B and C carry through an analysis parallel to that of the preceding example,
Jo
EJcosO
ElcosO
pL2
Note that this result is valid for an arbitrary variation of El. Finally, the total forces are
N1 N10 + Z1N1
N2
= N20 + Z1N21 =
M=M,0+Z1M,1 =0
Since M =
0, the deformed shape of the arch coincides with the initial shape when axial deformation is neglected. It follows that (c) also apply for the fixed nonshollolv case.
When the arch is shallow, the effect of axial deformation cannot be neglected. The
expression for Z1 follows from (1478):
SEC. 147.
471
Fig. E149A
p = Coflst
xi
Pnmary structure
R2,d2
Fig. E149B
N2
xI
L
N5,00
M0
pL(
Xj\
)
pL
472
CHAP. 14
dx1
pL2
8h
I
8h
1L1
+
jL1
consider A and I to be constant and evaluate for this geometry. The result is
I
8 Ah2
15(p'\2
8
Fig.
N11 = +1
N2,1 = 0
M,1 = +f
2x1
For the assumption of shallowness to be valid, 16(h/L)2 must he small with respect to unity. The total forces for the shallow case are
N1 =
pL2
M=
p1?
f(
\=
PL(
SEC. 148.
473
It is of interest to determine the rotation at B. The "Q" loading consists of a unit moment applied at B to the primary structure (see Fig. E149D). Applying (1477) (note that
Fig. E149D
PQ
+l
0),
we obtain
x1
P/
j
When El is constant, (i) reduces to
pL3(
2
(5
Since
0, the results for the fixed end shallow case will differ slightly from (h).
148.
One of the steps in the force method involves evaluating certain integrals which depend on the member geometry and the cross-sectional properties. Closed-form solutions can be obtained for only simple geometries, and one usually must resort to a numerical integration procedure. In what follows, we describe two proceduresi which can be conveniently automated and illustrate their application in deflection computations.
J
t See Ref. 8 for a more detailed treatment of numerical integration schemes.
(1479)
__________________
474
CHAP. 14
where f(x) is a reasonably smooth function in the interval XA divide the total interval into n equal segments, of length h:
We
h=
XA
(1480)
If f(x) is discontinuous, we work with subintervals and use a different spacing x,, represent the for each subinterval. For convenience, we let x0, x1, the coordinates of the equally spaced points on the x axis, and f0, corresponding values of the function. This notation is shown in Figure 1417.
.
.
12
1;,
a
XO
11
B
X,,
X1
X2
EXJk1,k =
f(x)dx
dx =
+
+
(1481)
f(x)dx
+ L)
(1482)
which is called the trapezoidal rule. A more accurate formula is obtained by approximating the curve connecting
fdx
J
+ 4fk+1 +
(1483)
Jk2= Jk +
To apply (1483), we must take an even number of segments, that is, n must
be an even integer. If the values of J at odd points are also desired, they can
SEC. 148.
475
be
determined using
h
=
Finally, one can express
in =
Jo
12
{Sfk +
Jk+2]
(1484)
as
+ 4(11 + .13 +
+
(1485)
+ 2(f2 +f4 +
Equation (1485) is called Simpson's rule.
N
fk1
fk
h
fk+1
S
Xk_1
Xk
fk
fk+1
fk2
S
Xk
Xkf.1
________________M
476
CHAP. 14
Example 1410
Consider the problem of determining the vertical displacement at Q for the straight member of Fig. E14lO. We suppose shear deformation is negligible. The deflection due
Fig. E1410
XQ
PQ
+1
XQ(1t)
MQ dx
(a)
where M is the actual moment and M0 is due to the "Q" loading. Substituting for M, expands to
/ 1L M
CXQ
'\
+
J
SEC. 148.
477
To evaluate (b), we divide the total length into ii equal segments of length h, number the points from 0 to n, and let
dx J0 El Cx M xdx El Jo
I
I
+ "k
Xkjk
If, in determining we also evaluate the integrals the interior points, then we can readily determine the displacement distribution using cd).
Example
1411
Consider the simply supported nonshallow arch shown. We suppose there is some distribution of It'! and we want to determine the vertical deflection at Q. Considering
Fig. E1411
ill
El
L
M,
ds
1W
1W,
dx
478
CHAP. 14
is the same as for the straight member, Then, the procedure followed in Example 1410 is also applicable here. We just have to replace M/EI with
M/EI cos 0 in Equation (c) of Example 1410.
REFERENCES
1.
TIMOSHENKO, S. 1.: Advanced Strength of Materials, Van Nostrand, New York, 1941.
2. 3.
REISSNER, E.: "Variational Considerations for Elastic Beams and Shells," J. Lag. Mech. Div., A.S.C.E, Vol. 88, No. EM 1, February 1962.
MARTIN, H. C.: introduction to Matrix lvi etliods of Structural Analysis, McGrawHill, New York, 1966. MUSRTARI, K, M., and K. Z. (IALIMOV: "Nonlinear Theory of Thin Elastic Shells,"
4.
5. 6. 7.
8.
Israel Program for Scientific Translations. Jerusalem, 1962 MARGUERRE, K.: "Zur Theoric der gekriimmten Platte grosser Formanderung," Proc. 5th mt. Congress App!. Mccli. pp. 93101. 1938. Onai'i, J. 1.: Mechanics of Elastic Structures. McGraw-Hill, New York, 1967. HILDEBRAND, F. J.: introduction to Numerical Analysis, McGraw-Hill, New York,
1956.
PROBLEMS
Specialize (147) for the case where Yi = x1. Let x2 = f(x1) and 141. let 9 be the angle from X1 to Y1 as shown below. Evaluate the various terms for a parabola
f=
144. 145.
on stress and displacement expansions (Equations (1425) and (1442)). Illustrate for the rectangular section treated in Example 14i. Which set of relations would you employ?
PROBLEMS
479
I
d
146.
2t
Prob.
142
T
b=O.75d
I
Evaluate I' and 1" for the symmetrical section shown. Prob.
146
h=O.75d
t=d120
147. Consider a circular sandwich member comprised of three layers, as shown. The core layer is soft (E 0). and the face thickness is small in comparison to the depth d). Establish force-deformation relations based
I I I
Starting with (1434) and (1435), derive a set of nonlinear strain 148. displacement relations for a thin member. Assume small finite rotation, and
480
CHAP. 14
=
Y12
e1
Determine the corresponding force-equilibrium equations with the principle of virtual displacements. 149. Refer to Fig. 1410 and Equation (1431). If we neglect transverse is orthogonal to t'1 and we can write shear deformation,
(1 +
tj
dt'1
ldF'
-
= fl1t1 + fl2t2
(a)
Verify that
can be expressed as
1
Cl
(i+e1
- y,k}
initial tangent vectors,
=
and take y
(b) S (i.e.,
1).
U1t1
+ U2t2
Derive the force-equilibrium equations, starting with the vector equations (see (1412) and Fig. 144),
dS
dM
+m +
x F = 0
and expanding the force vectors in terms of components referred to the deformed frame: = F11'I + = Mi3
(c)
h=
b11'1
Assume small strain. Derive the force-equilibrium equations with the principle of virtual displacements. Take the strain distribution according to Equation (b).
________
PROBLEMS
481
(d)
Derive the nonlinear deformation-displacement and equilibrium equations for the cartesian formulation. Refer the translations and loading to the basic frame, i.e., take
=
P
+ V212
Pi'i + P2t2
problem.
Specialize the equations for the case of a shallow member. 1410. The accompanying sketch applies to both phases of this
Prob. 1410
h2 = const
(a)
(b)
Determine the complete solution for the circular member shown. Utilize symmetry at point A = co = F2 = 0) and work with (1458), (1459). Discuss the effect of neglecting extensional and shear deformation, i.e., setting (1/A) (1/42) = 0. Repeat (a), using Mushtari's equations for a thin member with no
transverse shear deformation, which are developed in Example 142. Show that Mushtari's approximation (u1 << du2/dO) is valid when the segment is shallow. 1411. The sketch presents the information relevant to the problem:
Prob. 1411
P2 =
cOnSt
'j
L2
x2
482
CHAP. 14
shown. Consider the member to be thin and neglect transverse shear deformation. (b) Specialize (a) for negligible extensional deformation (set 1/A = 0).
(c)
Specialize (a) for the shallow case and investigate the validity of
Marquerre's approximation. 14-42. Refer to Example 146. Determine UB2 due to a uniform distributed constant. loading, b2 Determine the displacement measures at B (see sketch). Consider 1413. only bending deformation. Note: It may be more convenient to integrate the governing equations rather than apply (1469).
Prob. 1413
1414.
(a)
(b)
PROBLEMS
1415.
483
(a) (b)
Determine the radial displacement at B defined in Fig. E147. Determine the force solution for the loading shown.
Prob. 1415
P
1416.
x2
Determine the horizontal reaction at B due to the concentrated load. Consider the arch to be nonshallow. (h) Utilize the results of(a) to obtain the solution for a distributed loading
(a)
(c)
p2(x) per unit x1. Determine the reactions resulting from a uniform temperature increase,
T.
(d)
Suppose the horizontal support at B is replaced by a prismatic member extending from A to B. Assume the connections are frictionless hinges. Repeat parts a and c.
484
CHAP. 14
1417.
Prob.
1417
would generate the influence line for the horizontal reaction. Utilize the results contained in Examples 1410 and 1411.
15
In the first part of this chapter, we establish the governing equations for a member whose centroidal axis is an arbitrary space curve. The formulation is restricted to linear geometry and negligible warping and is referred to as the theory. Examples illustrating the application of the displacement and force methods are presented. Next, we outline a restrained warping formulation and apply it to a planar circular member. Lastly, we cast the force method for the engineering theory in matrix form and develop the member force-displacement relations which are required for the analysis of a system of member elements. The geometrical relations for a member are derived in Chapter 4. For convenience, we summarize the differentiation formulas here. Figure 151 shows the natural and local frames. They are related by
cos b
+ sin t'b
= sin4iii +
where = we obtain
= at
t
0
(15-1)
1t3
486
CHAP. 15
Fig. 15-i. Natural and local reference frames for a member element.
of a point, say Q, are taken as S and Y2, y3. The curvilinear Letting K be the position vector to Q (see Fig. 152),
R=
and differentiating, we find
13R
+ y2t2(S) +
(152)
= (t
y2a12
y3a23t2 + y2a23t3
(153)
= t2
3K
3y3
3K
=
where
(i
dS dy2 dy3
(154)
is the coordinate of Q in the normal (11) direction and the radius of curvature. Also,
3R 3R
..
= 1/K
is
(1
Y3
+
(155)
y2a23 =
/1
Y2
dcb
SEC.
487
0, which requires
a23 =0
d/
(156)
It is reasonable to neglect y/R terms with respect to unity when the member is thin, i.e., when the cross-sectional dimensions are small in comparison to
ay2
and
We express d4,/dS as
dq5
dS
(157)
where L is the total arc length and is the total increment in The nonorthogonality due to can be neglected when the member is only slightly
twisted, i.e., when
<<
1
(158)
where b is a typical cross-sectional dimension. In what follows, we will assume the member is thin, (158) is satisfied, and defines the orientation of the principal inertia directions.
488
CHAP. 15
Example 15i
The curvature and torsion for a circular helix are derived in Example 45:
1
R,
(H'\2
where R is the radius of the base circle and H is the rise in one full revolution. The helix is thin when b/R c< 1, where b is a typical cross-sectional dimension.
Example
152
By definition, a member is planar if r = 0 and the normal direction (Il) is an axis of symmetry for the cross section. We take the centroidal axis to be in the X1-X2 plane and define the sense oft2 according tot2 x 13 = Theangle is constant and equal to
either 0 (12
a13=a23=0
Example
153
is
S.
a13
(Ic!)
T=
If bk <
1,
const
IC
152.
FORCE-EQUILIBRIUM EQUATIONS
To establish the force-equilibrium equations, we consider the differential element shown in Fig. 153. We use the same notation as for the planar case. The vector equilibrium equations follow from the requirement that the resultant force and moment vectors must vanish:
dS
_. + rn
+ t1 x F =
We express the force and moment vectors in terms of components referred to the local frame, = FTt =
(1510)
m't
SEC. 152.
FORCE-EQUILIBRIUM EQUATIONS
bc/S
+
489
dS2
c/S
inc/S
+
c/S
where
F=
dF
tIS
dS
dS
Also,
x
= FJ3
=
=
{O,
_F3,F2}Tt
(b)
-a,
aF + b =
dM
+F0J
a12F2
a33F3 + b1 =
+
dF3
dM1
a12F1
a23F3
+ b2 = 0.
0 0
(1511)
a13F1
+ a23F2 + b3 = + m1 =
a12M2 a13M3
a12M1
a23M3 + m2 F3
When the member is planar, a13 = a23 = 0 and the equations uncouple naturally into two systems, one associated with in-plane loading (b1, b2, rn3,
490
CHAP. 15
F2, M3) and the other with out-of-plane loading (b3, in1, m2, F3, M1, M2). The in-plane equations coincide with (1414) when we set a12 = 1/R and the out-of-plane equations take the form
F1,
dF3
dM1 dM2
1
1.
in1
= =
(1512)
+ m2
F3
153.
We consider the material to be elastic and define V" as the complementary energy per unit arc length. Since we are neglecting warping restraint, is a function only of F and M. We let
Iav*
.
1, 2, 3
(15 13)
k = {kJ
and write the one-dimensional principle of virtual forces as
dS =
Js(eT
AF + kT AM)dS
(1514)
dS
It
(0
SEC. 153.
491
Now, we apply the principle of virtual forces to the element shown in Fig. 154. We define and ii5 as
= uTt = equivalent rigid body translation (1515) vector at the centroid = = equivalent rigid body rotation vector
The virtual system satisfies the equilibrium equations (155) identically and therefore is statically permissible. Evaluating APi,
AFT
an +
+ AMT
- aw) dS
10
an
+
(1516)
dO)
=
aY*
du1
e1 =
C2
du2
Cj2
a v*
+ a12u1 +
+ 023u2
e3 =
k1
(113
du3
aV*
dw1
=
DV"
=
dw2
0120)2
(1130)3
k2
k3 =
(*
+
+
0120)1
023C03
d0)3
0230)2
Once is specified, the left-hand terms can be expanded. The form. of depends on the material properties, the particular stress expansions selected, and the member geometry. In what follows, we consider the material to be linearly elastic and approximation with the complementary energy function
492
CHAP. 15
= +
where
+ +
+ + +
+
(15-17)
MT = M1 +
Y2,
= torsional moment with respect to the shear center coordinates of the shear center with respect to the centroid
F3y2
if
j-'-
dA
if
dA
=7+
M-,
M3
13
and using the shear stress distribution predicted by the engineering theory,
+ = is the where is the unrestrained torsional distribution due to MT and flexural distribution due to F2, In addition to these approximations, we are also neglecting the effect of curvature, i.e., we are considering the member to be thin. The approximate force -displacement relations for a linearly elastic thin curve member are
+
F2
F1
du1
=
e3
+
MT
=
=
du2
du3
+ a12u1
C21U3
F3
k1 =
MT
d1
M2
M3
UI3
18
a12co2 a13w3
dw2
a23co3
k3 =
dw3
U,)
+ a13co1 + 1223(02
SEC. 154.
493
When the member is planar, the shear center is on the Y2 axist and there is no coupling between in-plane (u1, U2, and out-of-plane (u3, (02) displacements. That is, an out-of-plane loading will produce only out-of-plane displacements. The approximate force-displacement relations for out-of-plane deformation for a thin planar member are
e3 =
F3
MT
du3
=
MT
dw1
I
+
(15-19)
k2
= k2
M2
= M1 y2F3. Note that flexure and twist are coupled, due to the curvature, even when the shear center coincides with the centroid.
where
154.
Since the displacement method involves integrating the governing differential equations, its application is restricted to simple geometries. In what follows, we apply the displacement method to a circular planar member subjected to out-of-plane loading. We suppose the cross section is constant and the shear center coincides with the centroid. It is convenient to take the polar angle e as the independent variable. The governing equations are summarized below
Rh3
M2 + R,n1
=0
dM2
+Rrn2RF3=O
F3
ldu3
=
=
M1
k1
I /dw1
+
M2
k2 =
o k2
l/dw2
t The shear center axis lies in the plane containing the centroidal axis, which, by definition, is a plane of symmetry for the cross Section.
494
CHAP. 15
Boundary Conditions
F3
M1
M2
The
or or or
u3
F3=C1RJ0b3d0
The remaining two equations can be transformed to
(1520)
(d)
dM1
+ Rin1
(e)
We solve (d) for M1 and determine M2 from (e). The resulting expressions are
+ Rm1
is
i313
F3
B
the X1 X2 plane.
are axes of symmetry.
SEC. 154.
495
The
+ w1 =
Rk2
R2
in1
(1 +
(f)
RM5
=
where
- Rw2
is a dimensionless parameter,
(1522)
which is an indicator for torsional deformation. Solving the first equation for and then determining W2 and u3 from the second and third equations lead
to
sin U + C5cosU +
RM1
C6
+j
dO
is the particular solution for wi. The complete solution involves six integration constants which are determined by enforcing the boundary conditions. The following examples illustrate the application of the above equations.
where
Example
154
The member shown is Sxed at A and subjected to a uniform distributed loading. Taking
Fig. E154
b3 = const
h3
F3 =
C1 Rb30
(a)
496
CHAP. 15
+
Then,
M1
= RF3 = RC1
R250
R2b30
= RF3 =
and the solution for M3 and M2 follows from (1521),
=
M2
C2
C2.sinO + C3cosO
F3=M1=M2=O
= C2 =
C3
R5308
R2h3 sin
at
0=08
R2b3 cos
Replacing
by
F3 = =
M2
Rb30
R2b3[q
sin
?q]
.R2h3[1 cos
Example 15-5
can be determined by applying the The force system due to the end action, librium conditions directly to the segment shown in Fig. E15SA. This leads to
F3 =
= FB3R(l
cos
F83R[l
0)]
(a)
M2 =
We
sin
sin(Oa
9)
suppose there is no initial deformation. Using (a), the equation for w3 becomes
+ w1 =
The particular solution of (b) is
(1
(1
0)
(b)
[0 cos(OB
9)]
Using the above results and specializing (1523) for this support condition lead to the following expressions for the displacements:
WAI COS 0 +
sin 0
C05
sin 0
cos(08
SEC. 154.
497
SlflO +
.
cosO +
.
+ cosO]
.
Sm 0
133
Osln(Oa
(d)
1) R5A2 Sifl
COS
0
SW
C1j sin P
1L
0
cos(05 0)
+ c, +
sln(OR
and define the rigid body displacements due to support movement. Also, terms involving c1 are due to twist deformation. The rotations and
Terms involving
Fig. E155A
F3
M2
A
I?
sin 0a
COS
R2P83 f
c1] sin
sin
+ h--
13)33
cos
033 -. 1]
1c,
+
+
C05
sin
Th
2c1
sin
El2
498
CHAP. 15
To investigate the relative importance of the various deformation terms, we consider the rectangular cross section shown in Fig. El 55)3. The properties aret
1
61 _6
(for
A3
5A 5d2d3
d3)
J =
Then,
(d3\2
El2
[1 (di'\21
2, 3 and v =
0.3
GA3R2
G [io k,i2) J
1,
for d3/d2 =
= E12/GJ
d3/d2
1
4k
1,69
(Ibr v
1.54
0.3)
2
3
2.75 3.16
3.8 7.4
Fig. E155B
1(3
t The torsional constant for a rectangular cross Section is developed in Sec. 113.
SEC. 155.
FORCE METHODEXAMPLES
499
Since (d2/R)2 << 1, we see that it is reasonable to neglect transverse shear deformation. In general, we cannot neglect twist deformation when the member is not shallow. For the shallow case, we can neglect in the expressions for UB3.
Example 156
Consider a closed circular ring (Fig. El 56) subjected to a uniformly distributed twisting 0 and M1, M2 are constant. Then, using (1516), we find
M1 =
1142
Rm1
RM2
WI
R2m1
=
x2
Fig. E156
b3 =
m2
const
155.
FORCE METHODEXAMPLES
In this section, we illustrate the application of the principle of virtual forces to curved members. The steps involved are the same as for the prismatic or planar case and therefore we will not reiterate them here. We restrict this discussion to the case where the material is linearly elastic, the member is thin
and slightly twisted, and warping is neglected. The general form of the expression for the displacement at an arbitrary point and the compatibility
equations corresponding to these restrictions (see (1514), (1517)) follow.
CHAP. 15
Displacement
at Point Q
= i
+ +
(k20
+ j [(ei
+
+
Fj,Q + (a-)
(1524)
Compatibility Equations
Z1, Z2
Zr = force redundants
= F30 +
= M3,0 +
R1 = R1,0 +
.fkJZJ
3=
1
kA
kZk
(k
1, 2
r)
(1525)
where
JMT.JMT.k +
+ ETM31M3,kldS + +
+
+
=
+
= M1 + y3F2
v2F3
= cos
+ sin
at
bt3
sin
+ cos
x1=O
SEC. 155.
FORCE METHODEXAMPLES
Now, we consider the problem of determining the translations of the centroid at B due to the loadingshown in Fig. E157A. It is convenient to work with translationcomponents (v52, v53) referred to the basic frame, i.e., the X2, X3 directions. We suppose that the shear
Fig. E157A
x2
x1
axts
center coincides with the centroid and cializing (1524), and noting that M1 = the displacement expression reduces to
dQ =
11L(i
+
(b)
Force Systems
The moment vectors acting on a positive cross section due to P2, P3 applied at B (Fig.
E157B) are
P2(L
=
To find M2, M3, we must These follow from Fig. E157C:
P3(L
x1)12
For P3.
M2 = P2(L
M3
==
P2(L
x1)cos
q5
502
CHAP. 15
Fig. E157B
1'2(Lx1)i3
'I
P3(L
L-x,
/B
P3
Fig. E157C
M2r2
P2(Lx1)z1
M3t3
For
M2 = M3 = +P3(L
1)cos 4)
x1)sin4)
(C )
Determination of
Due to P2
corresponds to P2 = +
1.
SEC. 155.
FORCE METHODEXAMPLES
503
Example 158
We rework Example 156 with the force method. Using symmetry, we see that
M1 =
=
Suppose the rotation w1 in the direction of in1 is desired. The virtual loading for this displacement is rn1 = + 1. Starting with
R2
COl
Example 159
Consider the closed ring shown. Only M1 and M2 arc finite for this loading. Also, the behavior is symmetrical with respect to X1 and we have to analyze only one half the ring.
Fig. E159
x2
42
44
504
CHAP. 15
f11Z1 =
A1
A1 = 2R
,,/2
[M1 0M1
1
+
dO
0M2,
El2
dO
f11
2R
= =
'
12
E12) sin
cos 0
0
[sin2
and it follows that Z1 = 0. We could have arrived at this result by noting that the behavior is also symmetrical with respect to X2. This requires M2 to be an even function of 0. The virtual-force system for WAI is T = + 1. Using (1524) and (a) leads to
2wAi = 2RJ
RTa[l
COAl
Example 15-10
We analyze the planar circular mcmber shown in Fig. El 5 bA. The loading is out-ofplane, and only F3, M1, and M2 are finite. To simplify the algebra, we consider the shear center to coincide with the centroid and neglect transverse shear deformation. It is convenient to take the reaction at B as the force redundant.
Fig. E151OA
A
P13
(Displacement restraint
iriX1 direction at B)
SEC. 155.
FORCE METHODEXAMPLES
505
Primary Structure
The primary structure is defined in Fig. El 5lOB:
B1 =
d1
R4
UA3
R2
MA2
= MAI
Z1
F53
d2 = d4 =
E151OB
M1 ,
xl
Force Analyses
The force solutions for the loadings shown in Fig. El 5 [OC are:
For F:
F'3,0 = +P
M1, = PR[l
M2,0 =
COS(ii
llc)]
(b)
PR sin(lJ 'Ic)
ForZ1 = +1:
'Ic
ii)
(c)
506
CHAP. 15
Fig. E151OC
=+i
B
fit
!ii =
R
M2
for
Substituting for the internal force and reactions, we obtain the following expressions and
R3 [(1 +
Sin
(1
Sill
COS
cos
+ R2
sin(05
O)dG
{o. [i +
cos(05
Or)]
sin
sin
Oc
+
sin
El2
GJ
SEC. 156.
507
15-6.
In what follows, we consider the member to be thin and slightly twisted. Referring to Fig. 152, these restrictions lead to
dR
d(vol.)
tl
dS dr2 dy3
(1526)
Therefore, in analyzing the strain at Q (S, y3), we can treat the differential line elements as if they were orthogonal. Thc approach followed for the prismatic case is also applicable here. One has only to work with stress and strain measures referred to the local frame 12, 13) rather than the global frame. Our formulation is based on Reissner's principle (1333):
d(surface area)] = r, = independent quantities = e(iI) p, b = prescribed forces = = complementary energy density
bTIi
V*)d(vol.)
We introduce expansions for fl, in terms of one-dimensional displacement and force measures (functions of S) and integrate over the cross section. The force-equilibrium equations follow from the stationary requirement with respect to displacement measures. We start with the strain measures, = One can show thatt Y12.
S1
ti
tj
where
au
+ t2
+
(3
-.
(3U
(1527)
is the displacement vector for Q (S, Yi' Y2). We use the same displacement expansion as for the prismatic case:
U1
IA2
f4i
(1528)
y3)
=
Expanding
co1(y2
Y3)
Y2)
=
t See Prob. 155.
+ cr13y13)dy2dy3
CHAP. 15
to
dy2
(i529)
The equilibrium equations consist of(1511) and the equation due to warping restraint,
MR
(1530)
which can be interpreted as the stress equilibrium equation for the weighted with respect to 4).
direction
Now, we use the stress expansion developed for the prismatic case. The
derivation is discussed in Sec. 135, so we only list the essential results here. The normal stress is expressed as
F1
=
where
M2
'2
Y3
M3 '3
+ -r
'1)
(1531)
the St. Venant warping function referred to the shear center. 4) = We write the transverse shear stress distribution as
+
=
are functions of
(1532)
+
+ +
M32)
+
+
(1533)
+ 2F2F3
+ +
JS(a124),2 + cr134),3)dy2dy3
MR = (1 +
A23
+ b2F2 + b3F3
br =
0
(1534)
SEC. 156.
509
and b2, b3 are due to self-equilibrating stress distributions.t reasonable, It is in this case, to take b2 0 and compute the shear coefficients (A2, A3) based on the primary fiexural shear stress distributions. Expanding the stationary requirement with respect to force measures yields the force-displacement relations,
ej
av*
k2
av*
k3
=
aF2
e2 + b2f =
aV*
e3 +
b3f =
i3V*
(1535)
, k3 are defined by (1516). The corresponding unrestrained where e1, e2, warping relations arc (1518).
.
Example
1511
To investigate the influence of warping restraint, we consider a planar circular member having a doubly svrnnietricai cross section (Fig. hisil), clampedat one end and subjected
Fig. E1511
NM
to a torsional moment at the other end. We neglect transverse shear deformation due to restrained torsion. The governing equations for this loading (See Sec. 154) follow.
Equilibrium Equations
dM1
= =
dM2
dO
M1
t See Prob. 156.
R dO + M'j
CHAP. 15
Force-Displacement Relations
M2 =
E12k2
El2
El4,
R
f
(10
f=
= GJk1
1 (dco1
(b)
Boundary ('onditions
0U
(01C02J0
1v11 =M
M2 =
One can write the equilibrium solution directly from the sketch:
M1
0)
M2 = Al sin(05
0)
M1 =
k1
GJk1
1
0)
(dco2
+ Wi)
'\
lvi
0)
and solve for k1, and then Wi. The resulting expressions are
2 2 =
G.J
.,.
El2
GJ
ME
0) cos
siiih
tanh
GJ +
=
{o cos(05 0)
sin 0 cos
cos(05 0)
sin 0 cos
{sinh
tanh
0
cosh
+ cos tanli
=
and
SEC. 157.
The rotation at B is
(RM\
l\ El2)
COS 0B)
K=
Sin 0B
cos2
tanh 704
If we set
On
On
and let (g) reduces to (1357), the prismatic solution. The influence of warping restraint depends on 2 and Values of K vs. 2 for = ir/4, it/2 are tabulated below:
1
for
for
1
= it/4
0.179 0.786 0.907
= ir/2
0.51)0
0.96
0.99
10
(open section)
(j)
2 =0
(closed section)
1
where t is the wall thickness and h is a depth measure. Since 2 = RA and R/li for a thin curved member, the influence of warping restraint is not as significant as for the
prismatic case.
157.
In the analysis of a member system, one needs the relations between the for
and displacements at the ends of the member. For a truss, these equations
512
CHAP. 15
to a single relation between the bar force and the elongation. Matrix notation is particularly convenient for this derivation so we start by expressing the principle of virtual forces and the complementary energy density in terms of generalized force and deformation matrices. Referring back to Sec. 153, we define
reduce
a v*
(1536)
aM1
dS
dS
dT
(1537)
Note that we are working with M1, not MT. We use the complementary energy function for a thin slightly twisted member with negligible warping restraint
(i.e., (1517)). With the above notation,
Eq. (1517)
where
g=
gfm
(15-38)
gf
Sym
A2G
Y2Y3
(7
I
GJ
o
J'3
ojo
1
El2
0
Sym
(1539)
We will use these general expressions for planar and out-of-plane deformation as well as for the arbitrary case. One has only to delete the rows and columns
SEC. 157.
513
(1540)
El
for planar loading applied to a planar member. Finally, we substitute fort in (15 37) and distinguish between prescribed and unknown displacements. The principle of virtual forces expands to
J5(t +
where
dS dT
dT
(1541)
contains prescribed displacements and R are the corresponding reactions; d contains unknown displacements and are forces corresponding to d. The virtual-force system must satisfy the force-equilibrium equations, (1511). It is more convenient to generate and R with the equilibrium equations for a finite segment rather than attempt to solve (15il). Consider the arbitrary member shown in Fig. 156. Each end is completely restrained against displacement. The positive sense of S is from A toward B.
13
514
CHAP. 15
suppose the geometry of the member is defined with respect to a basic frame which we refer to as frame ii, and take the end forces at B as the force redundants. Then, the primary structure consists of the member cantilevered
We
from A.
Throughout the remaining portion of the chapter, we will employ the notation for force and displacement transformations that is developed in Chapter 5. A superscript n is used to denote a quantity referred to (lie basic frame. When no frame superscript is used, it is understood the quantity is referred to the local frame. For example, represents the internal force matrix at point Q referred to the local frame at Q. Note that acts on the positive face. The force matrix for the negative face is The end forces at A, B are denoted and are related to the internal force matrices by by
=
=
15
42
= {u1,
U2, U3
{u}
(1543)
For this system, and are prescribed. We determine for the primary structure, i.e. the member cantilevered from A, due to displacement of A, temperature, loads applied along the member, and the end forces at B and then equate it to the actual The virtual-force
system is
Lw = AR = =
Also,
=
=
d=
d
+
Il
+
+
+
Next, we express
as
+
where
(1544)
is the internal force matrix at Q due to the prescribed external loading applied to the member cantilevered from A. Finally, substituting for
SEC. 157.
515
leads to
[SB
T(18 +
0)dS
(1545)
[JB
+ The first term is due to rigid body motion of the member about A whereas the second and third terms are due to deformation of the member. We define
as the member deformation matrix:
=
By definition, We also define
JSB
S-4
to rigid body
(1546)
notion aI,out .4
is
(1547)
fll
(1548)
Equation is the force-displacement relation for an arbitrary member with complete end restraint. It is analogous to the force-elongation relation for the ideal truss element that we developed in Chapter 6. The member flexibility matrix, f", is a natural property of the member since it depends only on the geometry and material properties. For simple members such as a prismatic member or a planar circular member with constant cross section, one can obtain the explicit form off. When the geometry is complex, one must generally resort to numerical integration such as described in Sec. 148 in order to determine f and This problem is discussed in the next section. Finally, we point out that the general definitions off, are also valid for in-plane or out-of-plane deformation of a planar member. One simply has to use the appropriate forms for the various matrices. Up to this point, we have considered only a simple member. Now suppose the actual member consists of a set of members rigidly connected to each other and the flexibility matrix for each member is known. We can obtain the total flexibility matrix by compounding the flexibility matrices for the individual elements. To illustrate the procedure, we consider two members, AA1 and
A1B, shown in Fig. 157.
The matrix, f", contains the displacements at B due to the end forces at B
with A fixed:
Now, suppose point A1 is fixed. Then, the displacement at B due to the tion of member A1B is
A1B
516
CHAP. 15
is the flexibility matrix for member A1B referred to frame n. The additional displacement at B due to movement of A1 is
where
B,' displaccment
at 4
' BA,
'Ai
It remains to determine
CF-n B
ffl
A,
ffl
AA,'
(TI?
It
Finally, we have =
+ =
= ffl
(1549)
1550
The first term is due to external load applied along the member and represents
The second and third terms are the end forces at B due to end displacement at B, A. Once is known, we can evaluate the interior force matrix at a point
using (1544),
(a)
Thus, the analysis of a completely restrained member reduces to a set of matrix multiplications once the member stiffness and initial deformation matrices are established.
SEC. 158.
517
sions for the end forces in terms of the end displacements are required. In
addition to (1550), we need an expression for
Now,
=
Substituting for
leads to
i
BA
A,0
B.i
represents the initial end forces. In order to express the equations where in a more compact form, we let
in KBB in
k AB
ii
inwn.T
n
BA
BA
jn
BA
BA
= =
Note that only
and
are
+ +
+
+
and
(1554)
15-8.
=
Noting that
=
and letting
g014)dS
A
=
ffl
(1555) (1556)
we can write
If numerical integration is used, the values of the integral at intermediate points along the centroidal axis as well as the total integral can he determined in the same operation. This is desirable since, as we shall show later, the intermediate values can be utilized to evaluate the initial deformation matrix. We consider next the initial deformation matrix:
=
We transform 4',g, and
0)dS
518
CHAP. 15
and
=
=
Suppose there is an external force system applied at an intermediate point, denote the force and moment matrices and the total force matrix:
say C. Let
(1559)
= Normally, the external force quantities are referred to the basic frame for the member, i.e., frame n. The initial force matrix at Q due to this loading is
given by
0
07fl Q.0
_II U
SA
Sc
(1560)
<C.
<C
Writing
or" or"
CD
Finally,
(1561)
=
With this notation, (f) simplifies to
(1562)
Also,
(1563)
The determination of the member flexibility matrix reduces to evaluating J defined by (1561). One can work with unpartitioned matrices, i.e., g, but it is more convenient to express the integrand in partitioned form. The partitioning is consistent with the partitioning of into F, M. Since the formulation is applicable for arbitrary deformation, it is desirable to maintain this generality when expanding in partitioned form. Therefore, we define a as the row order of F and /3 as the row order of M.
(1564)
Continuing, we partition
SEC. 158.
519
0
(1565)
=
(fixfj)
g11
()
gQ
() g12
E
i
g12
g22
The translation and rotation transformation matrices are developed in Secs. 51, 52 and the form of g for a thin curved member is given by (1538).
The local ilexihility matrix is defined by (1555). Using the above notation, the expressions for the submatrices are
=
(1566)
g22
g22 are diagonal matrices when the shear center Note that g12 = 0 and coincides with the centroid. If, in addition, axial and shear deformation are
neglected, g11 = We let
0.
=
The submatrices follow from (1565):
(1567)
+
'1112 '1122
XHQgZ2XBQ
(1568)
= =
-F-
'I' dS
I
Jp22j
x
(1569)
JP
= F"
ffl
is
-
ffl
(15-70)
520
CHAP. 15
The initial deformation matrix due to an arbitrary loading at point C can be determined with (1562). Its partitioned form is
=
1V01
r cli
rT
C,12
BC
C.12 Ii
(flxfl)
lYe)
where
L"c. 12 -
2V'BC
22J
'C
denote the initial translation and rotation matrices. We write The member stiffness matrix, is obtained by inverting
= (r)'
(
=I
[ku
12 22
k22j
I
(1572)
One can easily show that (we drop the frame superscript on
k,' k ii 11
1n
for convenience)
(1573)
\1 12)
22k $
If
T i
Once is known, the stiffness matriccs and can be generated. Expanding (1553) leads to the following partitioned forms:
knB
I
I
in KBA
rinIi
nvn,T llJtB,!
ml
12
I=
A
rnii
I
1n.!T 12
(1574
AA
[k';1
and presented a number of examples which illustrate the displacement and force methods of solution. Actually, we obtained the complete set of forcedisplacement relations and also the initial end forces for concentrated and uniform loading. Now, in this section, we generate the member flexibility matrix using the matrix formulation. We also list for future reference the various member stiffness matrices. The notation is summarized in Fig. 158. For convenience, we drop the frame reference superscript n, since the basic frame coincides with the local frame, i.e., = I. The positive sense of a displacement, external force, or end forces coincides with the positive sense of the corresponding coordinate
axis.
Starting with (1566), we have since R$ are identity matrices. = from (1568). Once XBQ is assembled we can determine the submatriees of
_________ ______
SEC. 159.
521
i2
MB!, WB i
UB!
I.
Now,
o
o 0
X5Q=
(LxQt)
(L.xQ1) 0
(a)
L/AE
/1 L
Sym
1)
+
+
\-2x3
I)
+
0
f12= GJ
LT2
0
.
V
2E13
+
(1575)
L2
GJ
2E12
0
0
L/GJ
=
Sym
10
LIEu
522
CHAP. 15
The submatrices of k are generated with (1 573), (1 5 74) and are listed below 03 0:
a2 =
2
12E12
a3
12E11
=
= I'---
'2
1 + a2
GJ +
12E
1+03
b1 =
AE
L
k11=
Sym
0
L3
0
0
6E1
L2
0 (1576)
k12 =
L2
k22
L2
0
(4 +
Sym
0
(4 +
0 0
Fl*
L
L3
L2
10
L2
L2
0
0
B=
1!
L2
SEC. 159.
MEMBER
b
MEMBER
523
6E!!Ye2
V
FI*
(4
L2
+ a3)_L_
Finally, the fixed end forces due to a concentrated transverse force and a
uniform transverse loading are summarized below.
concentrated Force
a3 =
12E13
MB3
FB2
(1577)
MBI =
MA1
+
F112
=
L
MA3 =
+ FB2 +
Force
a2
12E12 GA1L2
FB3
+
+ FB3)
(1578)
MAI
MB1
_PC3 FR3
MA2
MB2)
524
CHAP. 15
Concentrated Torque
MA1
= =
Tc1(1
(1579
F82 =
b2L
FA2
M83 = MA3 =
M81
h2L2
(1580)
MAI =
F83 = 1q43 =
M82 = =
Uniformly Distributed Torque, rn
M81
b3L.
b3L2
(1581)
= =0
MAI =
(1582)
1510.
In this section, we generate the flexibility and initial deformation matrices for a thin planar circular member, of constant cross section, using matrix operations. We include extensional and transverse shear deformation for the
sake of generality. Some of the relations have already been obtained as illustrative examples of the force and displacement methods. In particular, the reader should review Example 146, which treats planar deformation, and Examples
154, 155, 1510 for out-of-plane deformation.
The notation is summarized in Fig. 159. By definition, Y2 and Y3 are principal inertia axes and p3 = 0, i.e., the shear center lies in the plane containing the centroidal axis. It is convenient to take the basic frame (frame n) to be parallel to the local frame at B. The three-dimensional forms of
and
are
cos
sin
01
0
= =
= sin
=
cos
17]
= rRbq
0
1
(1583)
SEC. 15ID.
525
jO
I
R sin
R sin ,fl
for planar deformation and
R0=1
SQ
[R(1
cos
x;
Centroidal axis
b
B2
F;3,U;,
Mw
Fig. 159. Summary of notation for a planar circular member.
We
consider the member to be thin and use the local flexibility matrix
matrix.
526
CHAP. 15
ae
as =
Cs
El3
El7
El2
Ct
=
a1
ae + a7
a2 = ae
c1 =
(1584)
Y2\
\2
Y2
Symmetrica'
-
sin
+ az)i}
R3
- cosO5
+ a2)}
0
+r!j)
a2)sin
sin2 05(1
2
cos
+ c41
2c3
05 C2 Sifl
COS
R2
0
1,12
sin
T t212t)
cos05)
+
0
+ c2 sin
cos f15}
Ib 22
SEC. 15-10.
527
We consider next the determination of the initial deformation matrix due to an arbitrary concentrltted load at an interior point, C. Now, the flexibility matrix for the segment AC referred to the local frame at C, which we denote is known. We just have to change to and superscript b to c in by
When the external load is referred to the local frame at C, the displacement at C is given by
('IC
"B
BC
TrI &
T011hc.
BC
3
=
1, b .. V0 UB
=
U
j
r
T c
c
AC,11
v-I,. T
/1
Tc. T .4C.i21
\
c C
1'
+
Ii
(1585)
(
4C,121 C
Tcc
=
+
/
= M3
Slfl
+
1
1+a1
cos
'Ic
I'
1+02 Cos
0c --
+ -'-'--
R3 I
cog
1+a1
sin ic
o4
1:13
+
I
2
+ Sifl
R
5jfl
(1586)
s--L
Sill 'Ic
1+02.
+
2
j
+
00,3
WB3
sin
+ (1
COS
RO
528
CHAP. 15
Out-of-Plane Loading
V0.
R3 (
COS
'ic + c4)
C2
Sifl 0c cos 0B
+ c3( sin
sin
B + sin
6c COS 0J3 + C3 Sifl
)
+ + 1=
R21 R21
cos 'ic + C2
12 (
R2
CZS1flGCS1flOB
c3(1
j
COS 'ic + c3(sin 0B sin 'ic)
+ c2 sin
+
+
RI
cos
Cos 'ic
5jfl
(1587)
+ C2 Sifl c
o5}Tc2
2=
=
C2
12
COS
'ic)
SIfl Oc Sin
i-i
sin
E12
When the loading is symmetrical, one can utilize symmetry to determine the fixed end forces. The most convenient choice of unknowns is the internal forces at the midpoint, i.e., 6 = OB/2; F1 and M3 are unknown for the planar case and
only M2 is unknown for the out-of-plane case. Explicit expressions for the fixed end forces due to various loading conditions arc listed below.
Planar Loading
Fig. 1510 defines the notation for the planar case.
We consider two loadings: a concentrated radial force P applied at C, and a uniform distributed radial load b2 applied per unit arc length over the entire segment. The basic frame is chosen to utilize symmetry. We determine the axial force and moment at C from the symmetry conditions u1 = w3 = 0.
CASE 1CONCENTRATED RADIAL FORCE P
SEC. 1510.
529
1;.fl
B!
MA
SiflO!
CL
1+Q2
CL
(1 + a1\
CL
sin2 cc
cc
(1 + a2\
sin
cos
(1588)
'
131
RP ci -i7771
COS
Al
77
Cl
1'
'A2
MA
PR I
co.s
SIfl
+ 1,/I
(sin
COS
Rb2(1
sin
(1 + ai)
2
sin2
cc
cc
(1 + a2'\
sin cc cos
cc
530
CHAP. 15
Mc=R2b2aecb(_1
= =
C(
(1589)
a4)
cos
Rb2 Sill
= MA = R2b2aecb
Out-of-Plane Loading
Figure 1511 defines the notation for the out-of-plane case. We consider four loadings: a concentrated force P, and a couple Tboth applied at C; a uniform distributed force h3; and a uniform distributed couple
in1.
CD2
Mt
=0
PR
2
+ c3(l cosa)
SEC. 1511.
531
MB1 = MA1
--i-- (1
PR
cos
cx)
----- sin
PR.
cx
FB =
=
=
c2sin2cx
(1591)
FB=FA=0
CASE 3UNIFORM DISTRIBUTED LOAD h3
R 2 b3
c1(sincx
cxcoscx)
+
= R2b3(sin
cx
cx
C2
COS
= =
(1592)
cos cx)
1
= =FA=PRcx
R2b3(cx sin cx
+ cos cx)
= =
m1R-----
SJfl cx cos cx
(1593)
m1R sin cx
MA2 =
1511.
rn1R(1
cos
cx)
In this section, we develop the flexibility matrix for a member whose centroidal
axis is a circular helix. The notation is shown in Fig. 1512. The principal inertia direction, Y2, is considered to coincide with the normal direction, i.e., the inward radial direction, at each point. We also suppose the properties are constant. For convenience, we summarize the geometrical
532
CHAP. 15
relations:
x3 =
x2
R cos 0
R sin 0
a dO
x3=C8
dS
a=
=
fl
t2 = -COS
t3
sin 012
b=
---sinO
R.
cosO a
C -cos 0
IC
R
Rc, = R,3 =
sin 0
0
C(OB
C.
0)
R(sin
sin 0)
= C(OB 0)
R(cos
0
cos 0)
R(sin
sin 0)
R(cos
The steps involve only algebraic operations and integration. We first determine using (1566), then from (1568), and finally with (1570). In what follows, we assume the shear center coincides with the centroid and neglect
GJ
g2=
El3
reduce to
=
'I'12
'V22
SEC. 15li.
533
Centroidal axis
Yi
Notation-Din,ensionless Para,neters
=
C12
R2E12
R2
C2l2
C2 (E12\ 13
El2
R2 [12
ij T
El2
a3
RC[12
Ra3
a5
1+a1
a4+a6
2
a6
1a1
a7=
a9
a8=
a10
a6+3a4
2
a63a4
2
534
CHAP. 15
Elements of f71
[fit
= C2a(
+
R2a2cz 1
Sym
f22
1
I
f33j +
(1
02 B
OB}
El3
f
a6
+ Sin208)
a8
2sinO8 +
0B
COS
0B + a4(OB
05 + COS OB)}
El3
0B(08 sin
1 + cos
f31 = =
RCr1
+ (a1 a4)(i
cos0,1)
a4 sin2 05 + a9OssinOn}
cos
C2x
R2a2cc 1
+
/1
a10
2
sin 08 Cos
2a4OB
+El
= El2 + a8
f33 =
R2a
COS
sin
+
COS
0B + a4 sin
2a1
{(ai + a5)05
sin
a6 sin
cos
(1594)
Elements of
[114
Lf34
1161
f261
f35
f36j
COS
f14 =
f16 =
f24
El2
08}
El3
<08
sin
1 + cosoB}
+ a8 sin2 0B
<p8(08
a4(1 cos
Sffi 0B C05
SEC. 1512.
535
f26=
=
Ca3cJ
sin OB
a1(1
COS
=
f36 =
0B
cos
Elements
of
COS 013j
{a508 + a6 sin
Sym
a
{a508
a6
=
El2
sin 2
cos08)
aa3
112
1512.
In Sec. 157, we considered an arbitrary member which is completely restrained at both ends. This led to the definition of the member flexibility matrix and a set of equations relating the end forces and the end displacements. Now, when the member is only partially restrained, there is a reduction in the number of member force unknowns. For example, if there is no restraint against rotation at B, M8 = 0, and there are only a unknowns (where a is the order of F8), the rotation at B has no effect on the end forces. To handle the case of partial restraint, we first determine the compatibility equations corresponding to the reduced set of force unknowns. Inverting these equations and using the equilibrium relations for the end forces results in force-displacement relations which are consistent with the displacement releases.
Let Z denote the force redundants. Normally, one would work with the primary structure corresponding to Z = 0. However, suppose we first express the force at a point, say Q, in terms of the end forces at B, using, as a primary
structure, the member cantilevered from A:
(a
in terms
= EZ + G
(15-95)
536
CHAP. 15
0) due to the applied external if Z contains only end forces at B. Now, the principle of virtual forces requires
0
TO/fl
results in the compatibility equations for Z. It is convenient to work first with the virtual force system due to Equation (b) reduces to
T(,ffl +
T(O//,l
=
ffl are the initial deformation and flexibility matrices for thefull end using (1595), and requiring the resulting restraint case. Substituting for expression to he satisfied for arbitrary AZ, we obtain
where
(ETIftE)Z + ET(
(1596)
are the displacements of the supports at B, A. It should be noted that We suppose Z is of order q x 1, i.e., there are q force redundants. Also, we let t be the row order of (and 0?t).
(F
=
([3 x 1)
(1597)
E is i x q
Gisi x 1
and (1596) represents q equations. For convenience, we let
(q x q)
+ rG
(1
1)
(1598 )
trZ =
z)
Tcjjgn
ETJc)
'A
' O,Z
is the flexibility matrix for Z and it is positive definite since E must be of rank
q, i.e., the force systems corresponding to the redundants must be linearly independent. Note that one can determine directly by working with the primary system corresponding to Z = 0. This is the normal approach. The approach that we have followed is convenient when the member flexibility
matrix is known.
SEC. 1512.
537
At this point, we summarize the force-displacement relations for partial end restraint: Z = member force matrix
= EZ + G
f,.
0,z
ETrE
(15100)
f.Z
=
Note that, for complete end restraint,
'7
.
14'
G=0
member system. Continuing, we let
15101
kr =
The force redundants are obtained by inverting (1599):
(15102)
Z=
=
We defined
(15-103)
+G
(e)
15104)
= = =
+G
+ (I,
(15105)
+
ar"
'BA
(
A,0
iLl
I
+
I n'
RB
BA
'A
538
ENGINEERING THEORY OF AN
MEMBER
CHAP. 15
where
(15107)
=
k" AA
ar"
BA BA
BA e
T
BA
Comparing (15107) with (1553), the corresponding expressions for the complete restraint case, we see that one has only to replace by in the partitioned forms for and The equation for is different, however, due to the presence of the G term.
Example
1512
0. We take Z
- Fl +G
0.
=
and the effective stiffness matrix follows from (15104):
rc".
1
I
ii
,i
it
vn, T
it
by Er eliminates 9", the relative rotation at B. There is no compatibility requirement for the end rotations in this case; i.e., the support rotation at B, which we have defined as does not introduce any member deformation.
REFERENCES
I.
2.
3.
4.
5. 6.
REISSNER, F.: "Variational Considerations for Elastic Beams and Shells." J. Eng. Mech. Div. A.S.C.E., Vol. 88, No. EMI, February 1962. HALL, A. S. and R. W. W000HEAD: Frame Analysis, Wiley, New York, 1967. GEaR, J. M. and W. WEAVER: Analysis of Framed Structures, Van Nostrand, New York, 1965. RUEINSTRJN, M. F.: Matrix Computer Analysis of Structures, Prentice-Hall, 1966. LIVESLEY, R. K.: Matrix Methods of Structural Analysis, Pcrgamon Press, London,
1964.
DAEROWSKJ, R.: Gekriimmte dnnwandige Trger (Curved thin-walled beams), Springer-Verlag, Berlin, 1968.
PROBLEMS
7.
539
V. Z.: Thin Walled Elastic Beams, Israel Program for Scientific Translations, OfiIce of Technical Services, U.S. Dept. of Commerce, Washington, D.C., 1961.
BAZANT, Z. P.:
8.
PROBLEMS
151.
I/p j3\
/ \3E12J
and
x2
'2
b
x3
y3
13
153. Determine the reaction at B and translation (in the direction of F) at C for the member sketched. Neglect transverse shear deformation.
/
Shear center
Prob. 153
I Vertical restraint at B
540
CHAP. 15
154. Repeat Prob. 153, considering complete fixity at B. Utilize symmetry with respect to point C. Derive (1527). Start with the definitions for the strain measures 155.
(1 + r,)
=
Df)
D15
S[fl Y12 =
DS
E3f)
Dy2
DS Dy2
rectangular closed cell. Comment on whether one can neglect these terms. 157. Refer to Example 1511. Specialize the solution (Equations f) for = 1L 1. Verify that (g) reduces to the prismatic solution, (1357),
when
158.
*
0.
Consider a member comprising of three segments. Assuming the flexibility matrices for the segments are known, determine an expression for the member flexibility matrix in terms of the segmental flexibility matrices.
Generalize for n segments. 159. Discuss how you would apply the numerical integration schemes described in Sec. 148 to evaluate defined by (1569).
1510. 1511.
Determine the fixed end forces for the member shown, using (1577)
and (1579).
X2
Prob. 1511
1512. 1513.
Solve Prob. 153 using (1584) and (1587). Verify (1590) and (1591). Apply them to Prob. 154.
PROBLEMS
541
1514. Starting with (1587), develop expressions for the initial deformations due to an aribitrary distributed loading, h3 = b3(O). Specialize for b3 =
helix (constant cross section; Y2 coincides with the normal direction) developed
in Sec. 1511:
1)evelop a matrix equation for the displacements at B due to a loading referred to the global frame and applied at flint: See (1585). (b) Evaluate for the loading and geometry shown.
(a)
Prob. 1515
Y3, b
Y2,fl
I I
P
ir/2
c
=R/2 =E/2
x1 1516. Determine the reduced member flexibility matrix for no restraint against rotation at an interior point P.
1517.
Z =
MB M4}
Part IV ANALYSiS OF A
MEMBER SYSTEM
16
INTRODUCTION
We
respect to a global framet and use a superscript o to indicate quantities referred to the global frame. The external force and displacement matrices for joint k are denoted by
(p0)
(ax!)
(161)
(jOt)
(xxi)
(fbi)
where c is the number of translation (force) components, /1 is the number of rotation (moment) components, and i + fi. Note that = 2, /3 = 1 for a
planar system subjected to in-plane loading and 1, /3 = 2 for a planar /3 = 3. system subjected to out-of-plane loading. For an arbitrary system, In what follows, we assume the material is linearly elastic and the geometry is linear, i.e., we neglect the change in geometry due to deformation. The governing equations consist of joint force-equilibrium equations and member force-displacement relations. We have already developed the member forcedisplacement relations in Chapter 15, so that it remains only to establish the joint force-equilibrium equations. In this chapter, we apply the direct stiffness method, which consists in assembling the system stiffness and initial force matrices by superimposing the. contribution of each member. the next chapter, we present the general formulation for a linear member system and obtain the equations corresponding to the force and displacement solution by
t By global
frame, we mean a fixed Cartesian frame.
545
546
CHAP. 16
matrix operations. Finally, in Chapter 18, we extend the direct stiffness method
In developing the relations between the end forces and end displacements for a member, we considered the member geometry and loading to be referred to a basic member frame (frame n) and used A, B to denote the negative and positive ends of the member. The general relations were written (see (15107))
as
+ +
+ +
Note that (a) also applies when there is only partial end restraint or internal
releases.
Now, we define n_ as the joints at the positive and negative ends of member n. Replacing B by n, n the member frame take the form
+ +
\T
+ +
where
'
We transform the force and displacement quantities from the member frame to the global frame for the system by applying
= =
(162)
to (b). This step is necessary since we are working with joint forces and displacements referred to the global frame. The final expressions arc:
= =
+ +
+ +
(163)
where the global member stiffness and initial force matrices are generated with
k
TI
rzou
164
Once the displacements are known, we evaluate using (163) and then transform to the member frame. Since the initial end force and stiffness matrices are generated in partitioned form, it is natural to express (164) in partitioned form. Using the notation
SEC. 163.
547
plion
=
=
ip
t.MoJ
(flxl)
(165)
(Sxo)
()O,22
12
on,T
fi
00. 12
21
on on
on
P
k
1
21 =
0
on, Ti n
ii
11
o,T ( )( ), 12
(166)
"(')(),22
Don,
t 0,1
= depends on is a natural property of the member whereas Note that the orientation of the member frame with respect to the global frame. The
operations defined by (16--6) can be considered as the element matrix generation phase. The member force-displacement relations satisfy the equilibrium conditions
for the member and compatibility between the restrained end displacements and the corresponding joint displacements. Actually, the equilibrium condi= tions were used to determine Compatibility is satisfied by setting = When there is only partial restraint at an end, there will and be displacement discontinuities. For example, if there is a rotation release at the positive end, will not be equal to the end rotation matrix. We have treated'r partial end restraint by defining an effective member stiffness matrix k0. In the derivation of k0, we consider 1IA to be the displacements of the supports (i.e., the joints) and enforce continuity of only the restrained end
displacements.
163.
The equilibrium equations for joint k arc obtained by summing the end
548
CHAP. 16
depends on and the displacements of those joints which are In general, connected to joint k. We define as the total (or system) external joint force and joint displacement matrices:
167
=
where
91'o
(16-8)
contains the joint forces required to equilibrate the initial end forces We have dropped the reference frame superscript for convenience. The most efficient way to assemble and is to work with submatrices of order 1, the natural partition size, and superimpose the contributions of each member which follow directly from (163). This operation requires no matrix multiplications. The terms due to member n are listed below.
In
oT
inrown...,columnn_
Since
164.
is
joint translation restraints in the formulation for an ideal truss to an arbitrary member system. Actually, only the notation for the joint force and joint displacement matrices has to be changed. The governing equations are:
=
=
1
(1611)
The stiffness and initial force matrices are assembled using (169) and (1610). It remains to introduce the prescribed external forces and displaceis prescribed, and we just add ment restraints. If joint q is unrestrained,
SEC. 164.
549
is unknown. We replace q is completely restrained, to the matrix equation for with the matrix identity,
Finally, ifjoint q is partially restrained, some of the elements in 19q are unknown.
In this case, we replace the scalar equations for the unknown reactions by
scalar identities.
We suppose joint q is partially restrained and, for generality, consider the translation and rotation restraint directions to he arbitrarily orientated with respect to the basic frame. We define X'1, .. , as the orthogonal directions
.
for the rotational restraint frame. Quantities referred to the restraint frames are indicated with primes and a single superscript is used for the total matrix:
=
(flxj)
(1612)
L"i
Now,
P, = q
T"
/3
(1613
q
1
I
(1614)
L0
1615 )
in
PPO
The modification requires two operations. First, we transform (1611) to This is accomplished by premultiplying row q of
with
and postmultiplying column q of with 9/" In the second step, we replace the equations corresponding to the unknown elements in with identities. This operation can also be represented in matrix form. Suppose the rth element in is prescribed. We assemble four matrices, Eq, and as follows:
1.
EqandGq
E=I,
G=O,
+1
550
CHAP. 16
2.
We start with an ith-order column matrix having zero elements and set the element in row r equal to the prescribed displacement.
We
start with an ith-order column matrix having zero elements and enter
the values of the prescribed forces and moments referred to the restraint frames. Note that element r is zero. Premultiplying transformed row q of 3r, t?P0 with Eq reduces the rth equation to 0 = 0. Then, adding Gq to Eqirqq and to + q introduces the identity for the rth element and includes the prescribed external forces We also operate on the qth column of in to preserve syrnnwtry and include the terms due to prescribed displacements in The complete set of operations for joint q are listed below:
1.
=1,2,...,ql
T)a/e;
T)F
2.
9N, q
+
T)Eq + Gq
=
3.
(l616)
=q+i,q+2,...,j
= =
The operations defined by (1616) are carried out for each joint, working with successive joint members. We represent the modified equations as
(1617)
The superscript J is placed on % to indicate that the joint displacement matrices are referred to the local joint restraint frames, which may not coincide with the global frame. Again we point out that the primary advantage of this modifica-
tion procedure is that no row or column rearrangement is required. Solving (1617) yields the joint displacements (local restraint frame) listed in their natural order, i.e., according to increasing joint number. The modified stiffness
matrix,
Once QgJ is known, we transform the displacements from the restraint frames
to the global frame, using (1615), and evaluate the member end forces from (163). Next, we the total external force matrix, The contribution
SEC. 164.
551
of member n
in row fl...
(16IS)
Finally, we transform the external joint forces from the global frame to the local restraint frames. This step determines the reactions and also provides a statics check on the solution.
Example
Suppose
161
and
for E, G are
EqOi
and (1616) reduces to
Gqlj
,t,.
q
1.
=L2,...,ql
2.
3.
j
Oi
=
Example 162
Suppose
joint q is completely restrained against translation. Then, the translation are prescribed. The appropriate matrices for
r0
=
01
Gg
r1
=
Example 163
We consider the case where joint q is restrained with respect to translation in one direc-
tion and there is no restraint against rotation. This corresponds to a "roller" support. We take to coincide with the restraint direction and X'2, as mutually orthogonal directions comprising a right-handed system. The translation, is prescribed. The
prescribed forces are P52, and
552
CHAP. 16
We first assemble
From (1614),
=
where
rRoa'
I
[
r,s
13
1,2,3
=
The forms of E, G,
and
are
We specialize the results for a planar system subjected to planar loading. In order for only planar deformation to occur, the translation restraint direction must lie in the plane plane. It is convenient to select the orientation of the system, which we take as the of X'2 such that X3 coincides with X",. 'Ihe specialized forms are
=
= [S,.,]
0
Eq
r,s =
1,2
(d)
Gq
= ---H
0
Finally, we consider the case of a planar system subjected to an out-of-plane loading. direction in order for only The translational restraint direction must be parallel to the and arc prescribed. The out-of-plane deformation to occur. For this case, specialized forms are
Eq Cq
=
qe;
=
Note that (e) is obtained by settinge =
1,
REFEREt'JCES
REFERENCES
1.
2.
3.
MARTIN, H. C.: Introduction to Matrix Methods of StructuralAnalysis, McGraw-Hill, New York, 1965. RUBINSTEIN, M. F.: Matrix computer Analysis of Structures, Prentice-Hall, New York, 1966.
4.
17
We also establish variational principles for the force and displacement methods. Finally, we discuss how one can introduce member deformation constraints in the displacement method. Since the basic steps involved in the member system formulation are the same as for the ideal truss formulation, we recommend that
the reader review Chapters 6 through 9 before starting this chapter. Let r be the number of prescribed joint displacements. Then, the total number is of joint displacement unknowns,
na = if
(171)
The total number of force unknowns, flf, is equal to r (the reactions corresponding to the prescribed displacements) plus qT, the total number of member
force unknowns:
(172)
represents the number of force unknowns for member n. By definition, is equal to the number of force quantities that have to be specified in order to be able to determine the total internal force matrix at an arbitrary point. If the member is fully restrained at each end, q1, = i. For partial restraint, q,, is equal to i minus the number of independent force releases. Note that when the member is pinned at both ends, = 1 since there are only five independent moment releases.
554
SEC. 172.
MEMBER EQUATIONS
555
There are qT equations relating the member forces and the joint displacements.
Also, there are ij equilibrium equations relating the external joint forces and the member forces. The formulation is consistent, i.e., the number of equations
is equal to the number of unknowns. If flf if, the system is said to be statically determinate since the force unknowns can be determined using only the equi-
librium equations. The difference, flf ii, is generally called the degree of static indeterminacy, and represents the order of the final system of equations for the force method. For the displacement method, the final system ofequations arc of order In what follows, we first establish the member forcejoint displacement relations by generalizing the results of Sec. 1512. Then, we assemble the joint force-equilibrium equations. Finally, we introduce the joint displacement restraints.
172.
MEMBER EQUATIONS
The reduced member equations were developed in Sec. 1512. For convenience, we summarize the notation and equations below (see (15100)):
Z=
(ixi)
f"
member flexibility matrix (i x i) = ETfnE reduced member flexibility matrix (q,, x f, = member deformation matrix (i x i) = + f"G = initial member deformation matrix (i x i) = = )
These equations include the effect of partial end restraint, internal force releases,
and reductions due to symmetry or antisymmetry. We can also use (a) for complete end restraint by setting F = and G = 0. Now, we introduce new notation which is more convenient. First, we note that G contains the end forces at B due to the external member loads acting are the on the primary structure defined by Z = 0. Also,
end forces at A. Then we write
=G
(173)
is a compatibility Next, we note that the equation relating Z and K", requirement. The term fZ + ETeC?OZ is the relative deformation in the positive
sense of Z due to the member loads and the member redundants, Z, whereas is the relative deformation in the negative sense of Z due to support (joint) movement. The net relative deformation must be zero for continuity.
556
CHAP, 17
Then, we define
= reduced member deformation matrix (q,, x 1) = = reduced initial member deformation matrix x 1) = ETI/'O z = + f"G)
With this notation, the member equations take the form
174
R,o
(175)
+ frZ = We generalize the relations for member n by setting
Bn
A=n_
E=E,,
=
Since
Z=Z.
"r,
n
(17-6)
= fr.
the joint quantities are referred to the global frame, we must transform the end forces and displacements from the member frame (frame n) to the
global frame (frame o), using
=
The final equations follow.
Member ForcesEnd Forces
= =
TE)Z +
+
17 7
n+
17 ) 8
The force translation transformation matrix, 2C, is a second-order tensor, i.e., it transforms according tot
and it
T T
SEC. 173.
557
and 'r.
as.
TF)Z
.
since it is a natural property of the member whereas We prefer to work with depends on the selection of the global frame.
173.
force-deformation relations for member n. Equation (178) represents the By defining general flexibility and deformation matrices, we can express the member force-deformation relations as a single matrix complete set of equation. We let
x 1)
x 1) = total reduced member deformation matrix = 2 1, m} = total reduced initial member deformation matrix (q,. x 1)
ro,
ro, mJ
'1712
Note that f is quasi-diagonal, symmetrical, and positive definite. With this force-deformation relations are given by notation, the -V =
-V,,
+ fZ
(1713)
It remains to generalize the deformation-displacement relations. We define as the total joint displacement matrix referred to the global
frame.
(ij x 1)
=
and express 'V as
/4,
. .
(1714)
=
The partitioned form is
(1715)
d12
...
1
(17-16)
558
CHAP. 17
x i. Now, we see from (178) that there are only two non-zero elements in row n and they are at columns n_. The assembly of d is defined by
d
= sins = 0
S
17
17
fl_
1,
ii =
2,
. .
, in
_,
,..., 411,,,_ }
(im
x 1)
(mi x un)
E=
E
Em
(im x q1)
([718)
(zinxirn)
(un
x im)
= =
Next, we relate 411, 411_ to
ET9II(Q71+
T411)
(a)
the positive
C%
17
- 19
SEC. 174.
559
and column
=
and it follows that
17 20
d. =
=
=
where
174.
We have used the member force-equilibrium equations in developing the member force-displacement relations, so it remains only to satisfy equilibrium of the joints. There arc i equations for each joint, and a total of if equations. The expressions for the end forces in terms of the member forces are given by (177). Assembling the joint force.equilihrium equations involves only summing at each joint the end forces incident on the joint. as the total external joint force matrix referred to the global We define frame: (1722) =
1)
and
1'
(1723)
The elements of
members with Z
0.
are the joint forces due to external forces acting on the We express the complete set of equations as
+
Z1
p1,1
P1 =
+
I
Z2
Zm
(17-24)
We assemble
and
in row
CHAP. 17
Column n
TE
n_n
Tdyn
p171sn
= =
TE
(1726)
s=
fl+,
=
We let
(1727)
1'
1,,
2,
=
Then, we can express
as
(1728)
=
175.
+ C!1..O
(1729)
(17-30)
Now, we suppose r joint displacements are prescribed. We rearrange and that the prescribed displacements arc last. We also rearrange
-+
so
U =
P
cu1) (U2)
(rxl)
(1731)
=
Pf
(r
x
where U2, P1, and P1 are prescribed. We use B, A to represent the rearranged
forms
d:
(qr
I
x r)
B=
[liii [
_J
A2]
[An
[ A2
(1732)
fr'qr)
SEC. 175.
561
Pi =
(1733) (1734)
17 35
If the restraints are parallel to the directions of the global frame, the transformation of d to A (or to B) involve only a permutation of the columns of (rows The same permutation is applied to the rows Suppose joint q is partially restrained and the restraint directions do not coincide with the global frame directions. We first transform the force and dis-
placement matrices for joint q from the global frame to the restraint frame, using
A
O/Iq
=
This step involves postmultiplying column q of d by
T
(1736)
row q of
by
= J)1 + 4L
+ fZ =
(1737)
where the superscript J indicates that joint forces and displacements are referred
to local restraint frames. The final equations are obtained by permuting the the rows of and then partitioning. columns of d2 (rows of
The transformation of a71 to U can be expressed as a matrix product,
U = DQ/ =
where
(1738)
...
(1739)
and H is the row permutation matrix. One can generate H by starting with I and permuting the rows according to the new listing of the joint displacements, i.e., with the prescribed displacements last. Now, D is an orthogonal
matrix,
(1740)
Then,
= DTU
P=
562
CHAP. 17
P, A =
B
(1741)
AT =
D=
Finally, we can write
[D 1
LD2J
(n
x (j)
.,
(r
(1742)
A1 = A2 = P1 = 2 =
1
= BT
= Bf
(17-43)
B1Z =
P2
-.
which represents nj equations in unknowns. For the equations to he consistent for an arbitrary loading, the rank of 'B1 must equal n,,. Therefore, the stability requirement for the system is
r(B1) = ;(A1)
(1744)
Since B1 is of order
Equation (1744) is the stability requirement for a geometrically linear system. It is also the initial stability requirement for a geometrically nonlinear system.
In the next chapter, we develop the stability criteria for a geometrically nonlinear system subjected to a finite loading.
176.
NETWORK FORMULATION
In the formulation presented in the previous articles, we worked with the actual joint displacements and external joint forces referred to the global frame.
The governing equations are given by (1730), which we list below for
convenience:
=
where
Tc)
+
One assembles d, 9PI, using(1717), (1725), which are actually the expansions
of(b). By introducing new joint variables, we can express d in terms of only one
SEC. 176.
NETWORK FORMULATION
563
C. The rule (1717) for assembling d still applies except that now Let Y denote some arbitrary point. Suppose we express the actual force and
= statically equivalent force at Y due to the actual force matrix at joint k. (1746) = displacement at Y due to rigid body motion about joint k.
The actual and equivalent quantities are related by
Y, k
r)r
kY
(1747)
k
where
0
kY
I"
0 Xy2;
0
Xy1
0
4i)
plane
(17-48)
t planar
=
Substituting for we obtain
'Kr,,, =
(1748)
'Wy
(1749) (1750)
and write
'K =
The generation of
= =
S
= (q,, x 1)
(1751)
s=
1, 2
564
CHAP. 17
To express
=
+ Y
(irn
x irn)
(1752)
Then,
C_)
=
as
(1753)
We transform the joint forces, using (1747), and write the resulting equations
= =
+ fZ =
(17-54)
(if x i/)
jY
It follows that
,,
r
(1756)
g,YI =
The expression for
reduces to (17--53) when (d) and (c) are introduced. The formulation developed above can be interpreted as a network formulation since the connectivity term appears seperately in the factored form of d. A simplified version which does not allow for member force releases has been presented by Fenves and Branin (see Ref. 1). The only operational advantage of not working with the actual joint quantities is in the generation f d,m and d,,,. This advantage is trivial compared to the additional operations required ., to introduce the displacement restraints, and finally, to to generate to once the solution is obtained. Another serious disadvantage transform is that the equations tend to become ill-condii.ioned.
Fenves and Branin's primary objective was to show that the governing
equations for a member system can be cast in a form such that geometrical and topological effects are separated, i.e., a network formulation. DiMaggio and Spillars (Ref. 2) have also presented a network formulation for a rigid jointed member system. Actually their formulation is a special case of our first formulation. It is not, strictly speaking, a true network formulation since connectivity is not completely separated from geometry (see (1721)). The only way that one can separate connectivity from geometry is to redefine the joint variables. Note
SEC. 177.
DISPLACEMENT METHOD
565
that the ideal truss is an exception. Connectivity and geometry are naturally
uncoupled for this system. Whether one interprets the governing equations for a member system from a network viewpoint is of academic interest only. In the displacement method,
the equations reduce to the equations for the direct stiffness method. The only possible advantage of the network interpretation is in the force method. There one can use certain concepts of the mesh methodt to select a primary structure, provided that there are no member force releases or partial joint restraints. However, the selection of a primary structure for a rigid-jointed
frame having fixed supports is quite simple, and even this advantage is debatable.
177.
DISPLACEMENT METHOD
The governing equations are given by (1733), (1734), and (1735). Once the member forces are known, we can find the reactions from (1734). Now, we start by solving (1735) for Z in ternis of the displacements,
Z = Z1 + kA1UE + kA2U2
where
17-57)
= k=
(1758)
Z1, contains the initial member forces due to external loads acting on the
members and initial deformation resulting from fabrication errors or temperature changes. We substitute for Z in (1733) and write the result as
=
where
P0
+ K12U2
(nd x lid)
(1759)
= K12 = ATkA2 = +
x r)
(lid
X 1)
(1760)
The elements of P0 are the joint forces due to the initial end forces. Since A1 is of rank (when the system is stable) and k is positive definite, it
f See Sec. 95.
See Prob. 218.
566
CHAP. 17
that K1.1 is positive definite. Conversely, jfK1 is not positive definite, the system is unstable. The joint displacements are determined by solving (1759) and the member forces are obtained by back substitution in (1757). Operating on the restrained equations, as we have done above, is not efficient
since the various coefficient matrices must be generated by matrix multiplication. By first manipulating the unrestrained equations and then introducing
the displacement restraints, one can avoid any matrix multiplication. This
procedure corresponds to the direct stiffness method. Operating on (1730), we obtain
Z = Z1
+ kdll
(17-61)
(1762)
and
= =
+ c/TZ +
+
Equation (1762) is identical to (168). The of (169), (1610) when we introduce the factored forms of
d, Z1.
reduces to
ke.n
T n e.H
=
and
L0 .._1O fl+n+ e,fl
=
where
=
Finally, we expand (d):
(1763)
= +
T)C
One can easily show that (1764) reduces to (1610) when the properties of
C... are taken into account. The initial end actions for member n
=
T,
,,
SEC. 178.
FORCE METHOD
567
takes the
= +
The general form
(1765)
(1766)
In Sec. 164, we presented a procedure for introducing joint displacement restraints and represented the modified equations as
(ii eqs.)
Now, (f) consists of (1759) plus r relations for the prescribed displacements. We obtain (f) by starting with
[0
=
IrJ
L
U2
and permuting the rows and columns. This operation can be represented in
terms of the permutation matrix, 11, defined by
u=[IdllJ
= HTP
Then,
=
is positive definite when K11 is positive definite, i.e., when It follows that the system is stable.
178.
FORCE METHOD
B1Z =
P1
P1,
(fld eqs.)
(a)
BfU1 +
=
= 'V0 + fZ
2
+ 82Z
(b)
(c)
unknowns where Also, equations in Equation (a) represents The system is statically determinate when n4. We let B1 is of rank be the degree of static indeterminacy, i.e., the number of member force redundants: (1767) =
568
CHAP. 17
we can solve (a) for member forces in terms of the net prescribed joint forces (P1 P1 i) and r member forces. The compatibility
Since B1 is of rank
equations for the member force redundants are obtained by eliminating U1 from (b). This is possible since (b) represents qT equations whereas U1 is only x 1. In the next section, we specialize the principle of virtual forces of order for a member system and utilize it to establish the compatibility equations.
We suppose the first 0d columns of B1 are linearly independent. If the system is initially stable, the member force matrix Z can always be rearranged so that this condition is satisfied. We partition Z after row
Z
1)
1z,)
(ZR)
1)
(1768)
The elements of ZR are the force redundants for the system. We refer to the system obtained by setting ZR = 0 as the primary system. Continuing, we
partition B1 and B2 consistent with (1768):
(fld 0 qj-)
x fld)
("s
'Jo,)
B1
=[
(r
0
BIR
"d>
17
B2
(i
69
(r qo,)
P1
P1,2 +
We write the solution of (1770) as
ZP
BIRZR + B2RZR
(1770)
(1771)
KZR
(1772)
Zr,,, =
(B1pY'(P1
ZPR =
but it is not necessary to determine
Actually, the solution procedure can be completely automated.( The complete solution for is
xqo,)
z
Note that the member forces due to B1Z = 0. Finally, we substitute for
result as
(17-74) are self-equilibrating, i.e., they satisfy in the expression for P2 and write the
P2 =
t See Sec. 92.
P2,0
P2 RZR
(1775)
SEC. 178.
FORCE METHOD
569
where
(rXI)
=
(r X
P1,2 + J37PZP,0
B2R + B2pZp,ft
'1776)
It remains to determine Zft. Equation (b) represents qr equations in unknowns, U1. Since + = qft, there are excess equations. We partition (b) consistent with the partioning
of Z,
rDT 1
1
rDT 1
I.e'
2
*ftftJ o,R) L PR V R) and obtain the following two sets of equations relating to U1 and ZR:
L"2RJ
ft
BfPU1 +
'V'p
BfftU, +
+ = 1'O.J( +
+
+ fRRZR
(fld eqs.)
eqs.)
(1777)
(d)
The joint displacements can he determined from (1777) once Zft is known.
I/R + ZP +
+ fftftZft
+ fpftZR)
1778
Equation (1778) represents the compatibility equations for the force redun-
fZRZR = A
f
7T 4 7 1-q', R'PP'P. ft
'RR
7T c
o)
P2, ftU2
These equations are similar in form to the corresponding equations for the
ideal truss developed in Sec. 92. The flexibility matrix, fzR, can be expressed as
lil
[Zp
[ZP
ft
(1781)
Now, f is positive definite for a deformable system. Then, it follows that fzR is also positive definite. In a later article, we consider the case where certain member deformations may be prescribed. Once the preliminary force analyses have been carried out, the remaining steps are straightforward. We generate A, solve for Zft. and then determine
570
CHAP. 17
Zr,,
The final number of equations for the force method is usually smaller than for the displacement method VS. lid). However, the force method requires considerably more operations to generate the equations. The force method can be completely automated, but not as conveniently as the direct stiffness method. Also, automating the preliminary force analyses requires solving an additional set of nd equations. Another disadvantage of the force method is that the compatibility equations tend to he ill-conditioned unless one is careful in selecting force redundants.
179.
VARIATIONAL PRINCIPLES
In Chapter 7, we developed variational principles for the displacement and force formulations for an ideal truss, Now, in this section, we develop the corresponding variational principles for a member system. The extension is quite straightforward since the governing equations are almost identical in
form.
P = P, + ATZ
The partitioned form is
d"K = A AU =
A1
AU1 + A2 AU,
(17-82)
PTAU =
+ ZTd.K
(17-83)
be satisfied for arbitrary AU is equivalent to (a). If we consider U, to be prescribed, (1783) results in only (b). We refer to (1783) as the principle of virtual
displacements for a member system. In the displacement method, we substitute for Z in the joint force-equilibrium equations, using "1/',,) Z= = (AU
form of (1783) suggests that we define a scalar quantity, V = V(U), having the property dV = = dV(U) (1784)
The
t We work with the governing equations for the restrained system. See (1733), (1734), (1735).
SEC. 179.
VARIATIONAL PRINCIPLES
571
One
can interpret V as the strain energy function for the members. For the
V= =
17 85
as
V+
PTU
(17-86)
The Euler equations for H,, are the unpartitioned joint force-equilibrium
equations expressed in terms of U. Finally, we introduce the joint displacement constraint condition, U2 =02, by writing (1786) as
H,, =
1U1 +
U2)
(17-87)
where U1, U2, and P2 are variables. The Euler equations for (1787) are the partitioned equilibrium equations (Equations (h), (c)) expressed in terms of the displacements with U2 set equal to 02, i.e., they are the governing equations for the displacement formulation presented in Sec. 177.
=
where
1U1
(1788)
V=
form
+ A202
(1789)
The Euler equation for (1788) is (1759), and the second differential has the
'4
= AUrK11 AU1
(1790)
Since K11 is positive definite, we can state that the displacements defining the equilibrium position correspond to a minimum value of defined by (1788)
or (1787).
We consider next the force-method formulation. We let AP, AZ be a statically permissible virtual-force system. By definition,
= ATAZ = BAZ
(17-91)
Premultiplying both sides of (d) with AZT and introducing (1791) leads to the principle of virtual forces, APTU = (17-92)
Note that (1792) is valid only for a statically permissible virtual-force system,
i.e., one which satisfies (1791).
The compatibility equations follow directly from the principle of virtual forces by requiring the virtual-force system to be self-equilibrating. If AZ
satisfies
-
AP1
B1
AZ =
(17-93) (17-94)
572
CHAP. 17
[ZP,Rl
A ZR
(17-95)
AP2 =
We define the member complementary energy function, V's' = V*(Z), such that dV* = (1796) For the linear case,
and
= 4ZTfZ + ZT.K0
We also define the total complementary energy function,
as
(1797)
(1793)
The deformation compatibility equations, (1794), can he interpreted as the stationary requirement for 11. subject to the following constraints on Z, P2:
= =
+ B2Z
The constraint conditions are the joint force-equilibrium equations. Operating on (g), and noting that P, 2 are prescribed, lead to the constraint conditions on the force variations B1 AZ = AP2 =
self-equilibrating.
0 B2 AZ
Note that (h) require the virtual-force system to be statically permissible and
In the previous section, we expressed Z, P2 as
Z
P20 +
P2RZR
This representation satisfies (g) and (h) identically for arbitrary AZR, Sub-
+
ZR
ZR]
(17-99)
2. RU2 + const
The Euler equations for (1799) are (1779), and the second differential has the form
AZR
(17100)
SEC. 1710.
573
is positive definite, it follows that the true forces, i.e., the forces that satisfy compatibility as well as equilibrium, correspond to a minimum value
Since
of
Instead of developing separate principles for the displacements and force redundants, we could have started with a general variational principle whose Euler equations are the complete set of governing equations. One can easily show that the stationary requirement for
rIR = ZT(8TU1 +
T
Pfu1
IT
IT
710
considering
force-equilibrium equations and the member force-joint displacement relations. This principle is a specialized form of Rcissner's principle. We obtain (1787) from (17101) by introducing the force-displacement
=
and noting that, by definition,
V" = y
Suppose a member is assumed to he either completely or partially restrained with respect to deformation due to force. The rigidity assumption is introduced by setting the corresponding deformation parameters equal to zero in the local flexibility matrix, g. For example, if axial extension is to be neglected, we set 1/AE = 0. For complete rigidity, we set g Now, in what follows, we discuss the case where neglecting member deformation parameters causes the mmher flexibility matrix t. to be singular, This happens, for example, when axial extension is neglected for a straight member. The rank is decreasedt by I and the axial force-deformation relation degenerates to
=
where
= v,,
(a)
is the initial axial deformation due to temperature and fabrication error. Note that now the axial force has to be determined from the equilibrium equations. For rigidity, = 0, and the force-displacement relations
(see (175)) degenerate to
(b)
See (1675).
574
CHAP. 17
One can interpret (a), (b) as either member deformation constraints or as con-
straint conditions on the joint displacements. In general, the decrease in rank of the system flexibility matrix f is equal to the number of constraint conditions. We consider first the force method. The governing equations are given by
fZrZR = A
where
(qR eqs.)
=
"Zr
L
C.
Suppose these are c deformation constraints. Then, f is of rank order to solve (c), must be nonsingular, i.e., it must be of rank requires
qT
C
In This
(17402)
unconstrained member deformations. This condition is necessary but not sufficient as we will illustrate below. Aside from insuring that the flexibility matrix is of rank there is no difliculty involved in introducing member deformation constraints in the force formulation.
Example
171
q4 = 2
We take the forces in bars 3, 4 as the redundants:
cF1)
(F2J
IF3
ff4
Then,
ri Zr5[0
and
0
1
01
0
=
We can specify that, at most, two bars are rigid. No difficulty is encountered if only one bar is rigid. However, we cannot specify that bars 1, 3 or 2, 4 are rigid.
SEC. 1710.
0
relations, i.e., (1757) are not applicable. In what follows, we first develop the
appropriate equations by manipulating the original set of governing equations. We then show how the equations can be deduced from the variational principle for displacements. The governing equations are
P1 =
P1
+ AfZ
eqs.)
+ fZ = A1U1 + A2U2
(qT eqs.)
Now, we suppose there are c deformation constraints and the elements of are listed such that the last c elements are the prescribed deformations. We partition '/7' and Z as follows:
=
where
(cx1)
z
=
(17103)
contains the constrained member deformations and Z. the corresponding member forces. We use subscripts c, u to indicate quantities associated with the constrained and unconstrained deformations. Continuing, we partition
A1
(q-r
A
(cx I'd)
'ia)
(cxr)
)
(17104)
(qre)xi
-
(qT
xl)
05 (cx I)
fT
0
(cxc)
576
CHAP. 17
= 0. Note that, in order for f to be singular, there must be no coupling between and i.e., f must have the form shown above. Using this notation, the governing equations take the form
P1
= =
(17105)
(17106)
(17107)
Equation (17107) represents c constraint conditions on the unknown joint displacements, U1. The rank of A1, is equal to the number of independent constraint equations. One can easily demonstrate that c independent constraint
conditions are required in order to be able to analyze the system for an arbitrary loading.
Example
172
Suppose we specify that bars 1, 3 are rigid for the system considered in Example 17I.
= {e1, e1})
U21 = = u11 +
e3 =
U41
In what follows, we assume A1, is of rank c. We solve (17106) for 4 and substitute in (17105). This is permissible since is nonsingular. The resulting equations are
4=
=
and
(17108)
+ AfCZC = P1
(17109)
A1,U1 =
A2,02
(17110)
Now, the coefficient matrix, is nonsingular only when the structure obtained by deleting the restraint forces (4) is stable. By suitably redefining 4, we can transform (17109) such that the coefficient matrix is always nonsingular. Suppose we write
Z.
=4+ =4+
+ A2,02
(17ill)
where 4
represents
is an arbitrary symmetrical
SEC. 1710.
577
in (17109), we obtain
(a)
P1 =
P1,1
+
T
+
By defining
[k,,
1 ([A1,,
A2,,1
11
(17-112)
rk
L
rf
L
kcj =
Pi
+ ATk'(A2tY2
+ (Afk'A1)U1 +
= =
(17113)
(17-114)
(17115)
=
K11U1 +
= = =
K1202 = H1 = H2
(17116)
Since A1 must he of rank for stability and we have required to be positive definite, it follows that K1 is positive definite. Also, the solution for U1 must satisfy (17116) and we see from (17111) that is equal to the actual constraint force matrix, for arbitrary k'.
with Z. deleted, have the same form as the The expressions for Z and unconstrained expressions (1757) and (1759). Now, it is not necessary to
rearrange Z such that the constraint forces arc last. One can work with the
natural member force listing,
Z = {Z1, .
. .
, Z,,,}
and take arbitrary values for the member deformation parameters that are to be negelected. We obtain K11 and Fl1 by first generating using the
direct stiffness method and then deleting the rows and columns corresponding to the prescribed displacements. The constrained deformations, 1/v, can be listed arbitrarily. It is only necessary to specify the locations of the constraint forces (elements of in. the natural member force listing. Once the ments and constraint forces are known, we can determine the force matrix for
member n by first evaluating (see (178) and (1711))
Z =k'r,
fl
r,
/
(O/g '.
ro. ,,j
n
n -
"
n+
578
CHAP. 17
where k.,, is the modified stiffness matrix, and then adding the constraint forces in the appropriate locations. In what follows, we describe two procedures for
U1 =
and then substitute in (17116),
(17118)
112
(17119)
The coefficient matrix for is positive definite since K1 is positive definite is of rank c. Note that, with this procedure, we must invert an ndth and order matrix and also solve a set of c equations. For the unconstrained case, we have to solve only equations. Example
173
= U, =
To simplify the example, we consider only the effect of joint forces. Using the notation
{u1,u2}
Pz}
= e1
e2
4= 4=
F1
k,,
F2
= =
11,1=2
(P1
U2,
c=1
Fig. E173
0
Bar
is rigid
=
e2
1
SEC. 1710.
579
Ii
1]
Note that we cannot invert (17109), since Af,,k,,A1,, is singular. Now, we assume an arbitrary value for the stiffness of bar 2,
i:
=
L
kJ
LU
a
1
K11
= Afk'A1
k1[I'
=
K11U1 +
+
=
P1
[u-.]
0
(h)
(i) (j)
=
H1 =
The inverse of K11 is
H2 =
(k)
1 [1+2a ii
Then
(1k1[-l
+Ij
(1)
.L[1
=
and (17119) reduces to
ak1
+11
ak1
F'2p2p1
Substituting for F'2 in (17118), we obtain
U1
(n)
= 2p'
Id1
U2 = 0
U1, u2
in (h):
F1 =
F2
F'2
P2 P1
CHAP. 17
H2
which represents c relations between the displacements. Since A is of rank c, we can express c displacements in terms of rid c displacements, i.e., there are only nd c independent displacements. We suppose the first c columns of are linearly independent. Since is of rank c, we can always permute the columns such that this requirement is satisfied. We let
c
(17120)
and partition
U1:
(cx 1)
U1
(cxnd)
(Cxc)
(nxl)
(17121)
The elements of U are the independent displacements. By definition, is nonsingular. Then, solving (a) for the constrained displacements, we have
=
Finally, we express U1 as
2U
(17122)
U1 = BU + H3
where
a)
(17-123)
(cx
=
I
f fl
0
-J
(axx)
H3 =
(
(cxl)
j
0
(ax!)
= H1
=
(fld eqs.)
-- K11H3
114
from (b) by premultiplying by BT and noting (17125). The resulting system of n equations for U is
(BTK11B)U = BTH4
(n eqs.)
(17126)
Since B is of rank n, the coefficient matrix is positive definite. One can interpret
SEC. 1710.
581
BTK1
evaluate U1 from (17123). It remains to determine the restraint forces, We consider again Eq. (a). Assuming U.1 is known, we can write
K11U1
115
(n0 eqs.)
(17127)
The matrix, 115, is the difference between the external applied force, P1, and the joint force due to member force with the constraint forces deleted, i.e.,
115 =
where (see (17114))
P1,
1
ATZ
(c)
= k'(A1U1 + A2U2
(d)
We determine Z using the member force-displacement relations and assemble P1 + AfZ by the direct stiffness method. Now has c independent rows.
(17128)
H5
2
(17129)
Since
In this approach, we have to invert a matrix of order c and solve a system equations. Although the final number of equations is less than in the first approach, there is more preliminary computation (generation of B) and the procedure cannot be automated as easily.
of no c
Example
174
c4
The constraint conditions are
e1
U12
n=1
e10 e,0
e30
(a)
n42
e2
021
U22 032 031
144j
e3
e4
(b)
e4,
582
CHAP. 17
+1
1
+1
+1 +1
1-
U12(
1e10 +
Je,0
e30 + u32f
..e4,o + U41J
U31
I
U1
FL2
A1,
Columns 2,4, 5, and either I or 3 comprise a linearly independent set. Then, we can take either u1 or u21 as U. It is convenient to take U = u1 We permute the columns according to
1
1 *5
2.
3-.2
The rearranged form of U1 is
U1 = {u12,u21,u22u31 = (U, U}
u11}
We determine U,by applying (17122). This step is simple for this example since
I.
Finally, we assemble U1 defined by (e) and then permute the rows to obtain the initial
Fig. E174
2
x2
ears
+1
0
e1,,,
e2,0
-I
1142
U21
U22
= +1 {u11} +
0
0
1'
e30
+ 1132
u31
e4,0 + U4j
I
113
SEC. 1710.
583
The constraint forces are determined from (17127), which for this example has the form
+1 +1 +1
I
We
I
F4
=
H5,5
I H,
permute the rows of (g) according to (d) and consider only the first four equations. The resulting equations correspond to (17129).
It is of interest to derive the equations for the constrained case by suitably specializing the variational principle for displacements. We start with the
unconstrained form of
developed in Sec. 179,
=
where
V + !5L1U1
V=
"K = A1U1 + A202
=
Then, V reduces to
V=
+
)Tk('K
(17-430)
The elements of 4 are Lagrange multipliers. One can easily show that the stationary requirement for (17130) considering U1 and 4 as independent
variables leads to (17109) and (17110). Since = v", we can add the term
'(1"'
to (d). Taking
V=
(17131)
U1=BU+H3
(f)
584
CHAP. 17
=
V
V +
+ H3)
(17132)
= AUT[BT(P1 =
P1)
+
(g)
+ BTATkr(ASH3 + A2U2
BTH4]
Requiring to be stationary for arbitrary AU results in (.17126). Note that we could have used the reduced form for V, i.e., equation (d). Also, we still have to determine the constraint forces.
REFERENCES
1.
S. J., and F. H. BRANIN, JR., "Network-Topological Formulation of Structural Analysis," J. Structural Div., A.S.C.E., Vol. 89, No, ST4, August, 1963, pp.
FENVES,
483514.
2.
3.
DIMAGOT0, F. L., and W. R. SPILLARS. "Network Analysis of Structures," .1. Eng. Mech, Div., 4.S.C.E., Vol 91, No. EM3, June, 1965, pp. 169188. ARGYRIS, J. H;, "The Matrix Analysis of Structures with Cut-Outs and Modifications," Proc. Ninth International congress App!. Mech., Vol. , 1957, pp. 131142.
18
In this chapter, we extend the displacement formulation to include geometric nonlinearity. The derivation is restricted to small rotation, i.e., where squares of rotations are negligible with respect to unity. We also consider the material to be linearly elastic and the member to be prismatic. The first phase involves developing appropriate member force-displacement relations by integrating the governing equations derived in Sec. 139. We treat first planar deformation, since the equations for this case are easily integrated and it reveals the essential nonlinear effects. The three-dimensional problem is more formidable and one has to introduce numerous approximations in order to generate an explicit solution. We will briefly sketch out the solution strategy and then present a linearized solution applicable for doubly symmetric crosssections.
The direct stiffness method is employed to assemble the system equations. This phase is essentially the same as for the linear case. However, the governing equations are now nonlinear.
Next, we described two iterative procedures for solving a set of nonlinear algebraic equations, successive substitution and Newton-Raphson iteration. These methods are applied to the system equations and the appropriate rerelations are developed. Finally, we utilize the classical stability criterion to investigate the stability of an equilibrium position.
18-2.
Figure shows the initial and deformed positions of the member. The centroidal axis initially coincides with the X1 direction and X2 is an axis of symmetry for the cross section. We work with displacements (u1, u2, co3),
585
586
CHAP. 18
distributed external force (b2), and end forces (F1, F2. M3) referred to the
The governing equations follow from (1388). For convenience, we drop the subscript on x1, and M3, w3, 13. Also, we consider h1 = rn3 = 0.
Deformed position
b2 dxi
1182
x1 ,
Centroidai exis
Equilibrium Equations
=0
(F1u2,
+ F2) + b2
(a)
F2 =
Force-Displacement Relatio,is
F1
= ULx +
F2
2
= =
U2.
CO
(b)
(0,
SEC. 182.
587
Boundary Conditions
Forx =
0:
=
U2 =
UAI
w=
WA3
or or or
or or
1F110 = FAI
iF2 +
Mb =
FIlL
Forx =
U1
=
=
+FB1
IF2 +
Ml,.
P
F112
(d)
or
F1 =
F2 +
Pu2 = C3P
where C2, C3 are integration constants. We include the factor P so that the dimensions are consistent. The axial displacement is determined from the
first equation in (a),
PL
11111 UAI
('1.
(u2
dx
(182)
M = El
P\ u2
ri
Finally, the governing equation for u2 follows from the third equation in (e), +
where
+ C3)+
2__
El
=
co
(183)
cos px + C5 sin
+ C2x + C3 + U2b
sin j.tx + C5 cos px)
(184)
x
-
(i
+
C2
+ +
C4
b2 dx +
(i
b(EI
1/
(185)
588
CHAP. 18
0, L leads to four linear equations relating (C2 C5). When the coefficient matrix is singular, the member is said to have buckled. In what follows, we exclude member buckling. We also neglect transverse shear deformation since its effect is small for a homogeneous cross section. We consider the case where the end displacements are prescribed. The net displacements are
u = (u CD' = (a5
Evaluating (184) with A2 =
oc,
1.
(186)
U2bX)X_OL
we obtain
C2 = C3 =
1
jzC5
C4
C
1
sin 1iL
,u sin
.
(187)
C5 =
1tL
1 COS/LL
D=
buckling load:
2(1
Note that D 4 0 as
PJrnax
The end forces can be obtained with (ce). We omit the algebraic details
since they are obvious and list the final form below.
MA3 =
MB3
+
+
+ +
+ +
uA2)]
UA2)]
+
FB2 =
UA2)]
(u52 uA2)
(189)
2
C0A3
Unz)]
P1.
('Lj
(u2,
2
dx =
PL
erL
where
j
2(1 cos
D=
D62 =
pL sin pL
1iL cos
cos 1zL)
sin iL)
SEC. 182.
589
The
functions were introduced by Livesley (Ref. 7), and are plotted in Fig.
+ 27t. The initial end forces depend on 18 2. They degenerate rapidly as the transverse loading, b2. If b2 is constant,
bL
A2
52
bL2
1
(1810)
=
In order to evaluate the stiffness coefficients, P has to be known. If one end, say B, is unrestrained with respect to axial displacement, there is no difficulty since is now prescribed. The relative displacement is determined from
UB1 =
UA1
+
('L
PL
Le.
=
j
dx =
+
Dg54 =
C5
w43)
(UB2
L
(U52
sin
(1811)
=
= = =
(1
WAS) +
WA3
(1
eos
jiL'\
if
+
+ We call Cr the relative end shortening due to rotation. However, when both axial displacements are prescribed, we have to resort to iteration in order to evaluate P since e, is a nonlinear function of P. The simplest iterative scheme is
p(i+ 1)
(u51 UAI)
+
-
(1812)
for the incremental end forces due to increments in the end displacements are needed in the procedure and also for
stability analysis. If jiL is not close to 2ic, we can assume the stability functions
590
CHAP. 18
+02
pL
2
4 6
8
Oi
10
constant and equal to their values at the initial position, when operating
Aco,43 + c&2
AU42)]
(Au,32
dMB3 = dM3 +
dP42
+
+ AWA3
(1813)
h-- dP
.41
41
,JL'i
112 112
42
1 42
dFB1
dP
dF41 =
dP
dP =
Au41)
+ AEder
SEC. 183.
591
where
the incremental initial end forces are due to loading, Lxb2. We can
by assuming Au2
is constant.
AuA2)
(1814)
The coefficients in (1813) arc tangent stiffnesses. They are not exact since we have assumed and Au2, constant. To obtain the exact coefficients, we have to add
El3 F
1LA2
4)3] dCuL)
(18-15)
d(,uL) =
2b13
,
dP
- 41
=
ae
cIUA2
(1816)
pL)
der =
+ Au52 +
(WA
tie.
+ ---
+ --A(1zL)
in the equation for dP. An improvement on (1814) is obtained by operating on (1811), and assuming ,tL is constant.
183.
The positive sense of the end forces for the th case is shown in Fig. 183. Note that the force and displacement measures are referred to the fixed member frame. The governing equations for small rotations were
derived in Sec. 139. They are nonlinear, and one must resort to an approximate method such as the Galerkin scheme,t in which the displacement measures are expressed in terms of prescribed functions (of x) and parameters. The problem
is transformed into a set of nonlinear algebraic equations relating the eters. Some applications of this technique are presented in Ref. 5.
t This method is outlined in Sec. 1O6.
592
CHAP. 18
x2
M52, WB2
!152
//
//
x1
P2
Note:
The centroidal axis coincides with X1, X2 and X3 are principal inertia directions.
If we consider b1 = 0, the axial force F1 is constant along the member and the nonlinear terms involve and coupling terms such as co1M2;
Neglecting these terms results in linearized equations, called the Kap pus
equations. Their form is:
Equilibrium Equations
F1
P
+ x3w1,1) +
dx1
F2]+
b2
0
0
[P(u,3
+ F3] + b3 =
1+
1 + rn-1- +
d
dx1
1142
+ 711w1
1
=0
F3
= =
0 0
0
M, + F2 + in3 =
M4,1
(1818)
11
SEC. 183.
593
Force-Displacement Relations
u1,
+ u53,
CO3
+
"F2
F3
+
A23
w1,
X3r
I
U52,1
GA2
.. +
+
F3
A3
J
x2
U53,
+ Ui2 =
1/F2
GA23
(Dz,ij7f + C01,t
M2
M3
+ X3rF7 X2rF3)
for
x=
0)
+ F2 =
i) +
P(u,3
+ M!1 + M2 = M2
+ 711w1
M3 = M3
F3 = F3 = =
To interpret the linearization, we consider (1381). If one neglects the nonlinear terms in the shearing strains,
Y12
Y13
u12 + 112.1
u13 + U3,1
+
takes
u1
+ zfu2, i + U3,
and assumes
+ +
+
= = =
0 0
one obtains (1381). Equations are exact when the section is doubly symmetric, Assumptions (a) and (b) are reasonable if is small w.r. to u2 and However, they introduce considerable error when co1 is the dominant u3, term. This has been demonstrated by Black (Ref. 5). When the cross section is doubly symmetric,
1
=
1
= X2r = X3r =
A23
=0
(1819)
= r2
594
CHAP. 18
(r
is the radius of gyration with respect to the centroid) and the problem
We have already determined the solution for fiexure in the X1-X2 plane.
If we introduce a subscript for /L and
Cu2)2
P
El2
(1820)
=
and then replace
(02
U3 * U2
4
cbjCu3L)
F2
F3
F3
M'2
-
(1821)
F2
in (189), (1813), we obtain the member relations for flexure in the X1-X3
+
E12
U.42)]
(1822)
=
and
+
F42 =
U43)]
F43 =
[_WB2
W42
U43)]
U.43)
=P
P =
AE
Ci-'
J'Al =
+ AE(Cri + er2 + er3)
1
U41)
r2
CL
j
er3
dx1
er2
J
dx1
(1823)
2L
(u3
dx1
where is obtained from er2 by applying (1821). We generate the restrained torsion solution following the procedure described in Example 137. If the joints are moment resisting (i.e., rigid), it is reasonable
________1+P
SEC. 183.
_
595
r2P P= GJ
MB1
GJ
U
1+P
+ '>
Erlcb 1 +
<1)A1)
GJ
MB1
2
MA1
(1824)
1+
sinh 1uL
GJ(1 + P)
1 1coshuL [ + ---(1 sinh,uL /1(1 + Cr(1 + P))[ We neglect shear deformation due to restrained torsion by setting C,.
0.
0
I
c%)
+P
(1825)
At this point, we summarize the member force-displacement relations for a doubly symmetric cross section. For convenience, we introduce matrix notation:
= {F1P2F3M1M2M3}8
{u1u2u3w1w2w3}B
etc.
(1826)
=
where
+ kBJl%A + +
uA3);
0; 0;
contains
El3
El2
'P33
El3
El2 GJ
Sym
El3
596
CHAP. 18
AE L
El3
.
El2 GJ
El2
k44 =
El3
Operating on
leads
dimensional form of (1813). Assuming the stability functions are constant and taking
dP
UA2)(AU82
+
AWAI)}
iiA3)(AuB3
AuA3)
(1827)
+ r2(w81
we obtain
+ (knE +
+ (kBA
+ (knA
(18 28)
SEC. 18-4.
597
Pz
rp1
r 2 p1p3
r2p1p2
(r2p1)2
0
0 0 0
0
LFB1.
P2P3
kr
AE
L
Symmetrical
TAT
0
0
0
0 0
p3
uA2)
p2 =
UA3)
Pi =
WA1)
1fF is close to the member buckling load, one mustinclude additional terms
due to the variation in the stability functions and use the exact expression for der
Kappas's equations have also been solved explicitly for a monosymmetric section with warping and shear deformation neglected. Since the equations are linear, one can write down the general solution for an arbitrary cross section. It will involve twelve integration constants which are evaluated by enforcing the
In this section, we present the mathematical background for two solution techniques, successive substitution and Newton-Raphson iteration, and then apply them to the governing equations for a nonlinear member system. Consider the problem of solving the nonlinear equation
=
Let
(1829)
x=
g(x)
(1831)
(1832)
where
x=g
t See Ref. 9.
598
CHAP. 18
Then,
=
in a Taylor series about
Expanding
g(k)
+ g(T
and retaining only the first two terms lead to the convergence measure
(1833)
where
is between
and T.
Ax +
Ax =
=0
=
+
(1834)
The convergence measure for this method can be obtained by combining (a)
and (1834), and has the form
(1835)
Note that the Newton-Raphson method has second-order convergence whereas successive substitution has only first-order convergence. We consider next a set of n nonlinear equations:
'I' =
=
An exact solution is denoted by In successive substitution,
x2,...,
Also,
(1836)
= is rearranged to
c g
ax =
(1837)
(1838)
a5 =
Then,
1))
g(k))
t See Ref. 9.
SEC. 184.
599
g(k)
+ =
+
91,2
92,2
'
9i,i
=
92,1
91n
92,n
L0xrJ
and retaining only the first two terms results in the convergence measure (x
(1839)
'g. must be
where
= = =
= [T'j
(1840)
= =
( 1841 )
(1842)
Let us now apply these solution techniques to the structural problem. The governing equations are the nodal equations referred to the
qlobcil system frame,
1?e
=0
(1843)
contains the external nodal forces and is the summation of the member end forces incident on node i. One first has to rotate the member end forces, (1826), from the member frame to the global frame using =
where
k =
is constant. We introduce the displacement restraints and write the final equations as
Pm
P1
+ KU
1844
600
CHAP. 18
depend on the axial forces while Pr depends on both the axial force and the member rigid body chord rotation, if the axial forces are small in comparison to the member buckling loads, we can replace K with
Note that K and
KU =
1.
K1.
2.
(ni_pe(1)
_p(n1)
e(2)
r
3.
N=
'\
1
Jabs
e
va]uc
(a specified value)
(1846)
This scheme is particularly efficient when the member axial forces are small
with respect to the Euler loads since, in this case, we can take K = K1 during the entire solution phase. In the Newton-Raphson procedure, we operate on Vi according to (1841):
=
4,(fl)
due to
+ dPr + K LW +
(1847)
=
where
= =
1848 )
iterate on (1848) for successive load increments. This scheme is more has to be updated for each cycle. However, its convergence rate is more rapid than direct substitution. If we assume the stability functions
We
expensive since
SEC. 184.
601
are
constant in forming
dP,
K + Kr
(1849)
where K. is generated with (1828). We include the incremental member loads at the start of the iteration cycle. Rather than update at each cycle, in fixed for a limited number of cycles. This is called mod (fled one can hold
Newton-Raphson. The convergence rate is lower than for regular NewtonRaphson but higher than successive substitution.
We consider next the question of stability. According to the classical stability criterion,t an equilibrium position is classified as: stable neutral unstable
>0
0
(1850)
d2
d2W,,, d2We
d2 W,,
<0
is the second-order work done by the external forces during a where d2 displacement increment AU, and is the second-order work done by the member end forces acting on the members. With our notation,
d2w,
Pe)TAU
=
(d
AU
(1851)
= AUTK, AU
\T
< 0
Pe) AU = 0 >0
(1852)
The most frequent case is Pe prescribed, and for a constant loading, the tangent stiffness matrix must be posil.ive definite,
To detect instability, we keep track of the sign of the determinant of the tangent stiffness matrix during the iteration. The sign is obtained at no cost (i.e., no additional computation) if Gauss elimination or the factor method are used to solve the correction equation, (1848). When the determinant changes sign, we have passed through a stability transition. Another indication of the existence of a bifurcation point (K1 singular) is the degeneration of the convergence rate for Newton-Raphson. The correction tends to diverge and oscillate in sign and one has to employ a higher iterative scheme. Finally, we consider the special case where the loading does not produce significant chord rotation. A typical example is shown in Fig. 184. Both the
t See Sees. 76 and 106.
______________________
602
CHAP. 18
frame and loading are symmetrical and the displacement is due only to short-
ening of the columns. To investigate the stability of this structure, we deletet the rotation terms in K, and write
K
The member axial K is due to a unit value of the load parameter forces are determined from a linear analysis. Then, the bifurcation problem reduces to determining the value of 2 for which a nontrivial solution of
(K + 2K;)AU
0
(1854)
Fig. 184. Example of structure and loading for which linearized stability analysis is applicable.
K, AU =
2K AU
(18-55)
Both K, and K; are symmetrical. Also, K1 is positive definite, Usually, only the lowest critical load is of interest, and this can be obtained by applying inverse iterations to
(K;)Au
1
(1856)
REFERENCES
1.
2.
3.
M. GERE: Theory of Elastic Stability, 2d ed., McGraw-Hill, New York, 1961. KOLLBRUNNSR, C. F., and M. MEIsTER: Knicken, Biegedriilknicken, Kippen. 2d ed., Springer-Verlag, Berlin, 1961. BLEICH, F.: Buckling Strength of Metal Structures, McGraw-Hill, New York, 1952.
REFERENCES
4:
BLYRGERMEISTEa,
603
5. 6. 7.
G.. and F!. STEUP: Stabilidhsrheorie, Part 1, Akademie-Verlag. Berlin, 1957, CFJtLVER, A. H., ed.: Thin- Walled Structures, Chatto & Windus, London, 1967. VLASOV, V. Z.: Thin Walled Elastic Beams, Israel for Scientific Translations, Office of Technical Services: U.S. Dept. of Commerce, Washington, D.C., 1961. LIVESLItY, R. K.: Matrix Methods of Structural Anal vsis, Pergamon Press, London,
1964.
8. 9.
10.
AROYRIS, J. H.: Recent Advances in Matrix Methods of Structural Analysis, Pergamon Press, London, 1964. HILDEBRAND, F. B.: Introduction to Numerical Analysis, McGraw-Hill, New York,
1956.
11.
GALAMBOS, T. V.: Structural Members and Frames, Prentice Hall, 1968. BRUSU, D. and B. ALMROTH: Buckling of Bars, Plates, and Shells, McGraw-Hill, New York, 1975.
index
Associative multiplication, 8
incidence ma-
grange multipliers, 76, 80 Curved member definition of thin and thick, 434 thin, 487 slightly twisted, 487
planar member,
stretching
and
Center of twist, 383, 389 Characteristic values of a matrix, 46 Chord rotation. p. 586 Circular helix, definition equation, 84, 86
Circular segment out-of-plane loading, 504 restrained warping solution, 509 Classical stability criterion continuum, 256 member system, 603 truss, 170 Closed ring, out-of-plane loading, 503 Cofactor, 19 Column matrix, 4
Degree of statical indeterminacy member, 555, 567 truss, 210 Determinant, 16, 37, 39 Diagonal matrix, 10 Differential notation for a function, 70, 72,
79
Direction cosine matrix for a bar, 119 Discriminant, 40, 59 Distributive multiplication, 8
Echelon matrix, 29 Effective shear area, cross-sectional properties, 302 Elastic behavior, 125, 248 End shortening due to geometrically nonlinear behavior, 589
Column vector, 4
Complementary energy continuum, 261 member system, 572 planar curved member, 434
606
INDEX
First law of thermodynamics, 248 Fixed end forces prismatic member, 523 thin planar circular member, 528 Flexibility matrix arbitrary curved member, 515 circular helix, 534 planar member, 462 prismatic member, 345, 521 thin planar circular member, 526 Flexural warping functions, 296, 300/n Frenet equations, 91
Member, definition, 271 Member buckling, 588 Member force displacement relations. 537, 546, 556 Member on an elastic foundation, 384, 369 Mesh, network, 220
Gauss's integration by parts formula. 254 Geometric compatibility equation arbitrary member, 499 continuum, 259, 264 member system, 569 planar member, 463, 466 prismatic member, 355 truss, 160, 212, 216, 223 unrestrained torsion, 279, 315 Geometric stiffness matrix for a bar, 200 Geometrically nonlinear restrained torsion solution, 595 Green's strain tensor, 234
Hookean material, 126, 249 Hyperelastic material, 248
Incremental system stiffness matrix member system, 601 truss, 193 Inelastic behavior, 125 Initial stability member system, 562 truss, 137 Invariants of a matrix, 59, 62 Isotropic material, 252
Kappus equations, 592 Kronecker delta notation, 11
Natural member reference frame, 92 Negative definite, 58 Negligible transverse shear deformation, planar member, 443, 454, 498 Network, topological, 220 Neutral equilibrium, 170, 256, 601 Newton-Raphson iteration, member system, 598 Normalization of a vector, 49 Null matrix, 4
One-dimensional deformation 335, 338, 432
measures,
Permutation matrix, 42, 135 Permutation of a set of integers, 16, 37 Piecewise linear material, 126, 146 Plane curve, 98, 425 Poisson's ratio, 252 Positive definite matrix, 58, 63 Positive semi-definite matrix, 58 Postmultiplication, matrix, S Potential energy function, member system,
571
Lagrange multipliers, 76, 80, 583 Lagrangian strain, 234 Lam constants, 253 Laplace expansion for a determinant, 20,
38
Linear connected graph, 218 Linear geometry, 120, 143, 237 Linearized stability analysis, 602 Local member reference frame, 92
Marguerre equations, 449, 456 Material compliance matrix, 249 Material rigidity matrix, 249 Matrix iteration, computational method,
201
Premuftiplication, matrix, 8 Primary structure member system, 568 planar member, 463 prismatic member, 354 truss, 211 Principle minors, 55 Principle of virtual displacements member system, 570 planar member, 442 Principle of virtual forces arbitrary member, 490, 492, 512 member systens, 571 planar member, 435, 458 prismatic member, 338, 351
Quadratic forms, 57
INDEX
607
Radius of gyration, 434 Rank of a matrix, 27, 42, 43 Rayleigh's quotient, 75, 79 Reissner's principle continuum, 270 member, 383, 414 member system, 573 Relative minimum or maximum value of a function, relative extrema, 66
Restrained torsion solution, member linear geometry, 391 nonlinear geometry, 595 prismatic
prismatic member, geometrically nonlinear behavior, 588, 595 prismatic member, linear geometry, 522 Strain and complementary energy for pure
torsion, 280 Strain energy density. 248
Stress and strain component trnasformations. 249 Stress components Eulerian, 242 Kirchhoff, 246 Stress function, torsion, 276 Stress resultants and stress couples, 272 Stress vector, 240 Stress vector transformation, 242 Submatrices (matrix partitioning), 12, 36 Successive substitution, iterative method member system, 597 truss, 193 Summary of system equations, force equiibrium and force displacement, 561 Symmetrical matrix, Il, 35 System stiffness matrix member system, 548, 550, 565 truss, 179, 180, 188, 206
Rotation transformation matrix, 101, 232 Row matrix, 4 Self-equilibrating force systems, 160, 211,
258
member systems, 568 Shallow member, assumptions, 448 Shear center, 297, 300, 309, 378, 389 Shear flow, 287 Shear flow distribution for unrestrained torsiOn, 308 Similarity transformation, 53, 62 Simpson's rule, 475 Singular matrix, 22 Skew symmetrical matrix, 11 Small strain, 120, 235 Small-finite rotation approximation, 238 Square matrix, 4
Tangent stiffness matrix for a bar, 193 prismatic member, 590, 596 Tensor invariants, 232 Torsion solution, rectangle, 281 Torsional constant, J, 276, 278, 323 Torsional warping function, 274, 377 Transverse orthotropic material, 252 Transverse shear deformation planar member, 454, 498 prismatic member, 355
irapezoidal rule, 474 Tree, network, 220 Triangular matrix, 12 Two-hinged arch solutions, 467, 470
Unit matrix, 10
Variable warping parameter, f, for restrained torsion, 372 Vector, definition (mechanics), 4/n
Work done by a force, definition, 153, 156