Applications of Energy Methods: - Principle of Stationary Potential Energy - Castigliano's Theorem
Applications of Energy Methods: - Principle of Stationary Potential Energy - Castigliano's Theorem
Applications of Energy Methods: - Principle of Stationary Potential Energy - Castigliano's Theorem
Widely used to obtain solutions to elasticity problems Basis for the Finite Element Method Energy is a scalar quantity Can be used to obtain elastic deflections of statically indeterminate structures and to determine redundant reactions Two Principles will be presented:
Principle of Stationary Potential Energy Castiglianos Theorem
1
(a)
where (Q1, Q2, , Qi, , Qn) are the components of the generalized load If dxi is a translation Qi is a force If dxi is a rotation Qi is a moment
2
d W = d We + d Wi
where
(b)
d We is the virtual work of the external forces d Wi is the virtual work of the internal forces
The virtual work done by the external loads can be expressed as
(c)
For an elastic system where the loading is path independent, the virtual work of the internal forces is equal to the negative of the virtual change in the elastic strain energy, d U
d Wi = - d U
(d)
where U=U(x1, x2, , xn) is the total strain energy in the system For a conservative system, dW = 0 which is known as the Principal of Stationary Potential Energy
3
(e)
But the dxi can assume any value, so for the above equation to be true requires
(f)
For any system with finite DOF, if the components Qi vanish, then the system is in equilibrium a system is in equilibrium if
For a simple example of this theorem, consider a uniform bar loaded at its end by a load P
Let the bar be made of a nonlinear elastic material The area below the curve represents the total strain energy U stored in the bar, By Eq. 5.1, P = U/ e If the load-elongation data are plotted as s-e, then and s = U0 /e (as given by Eq. 3.11)
See nice Example 5.1 in Boresi text Fig. 5.1
5
5.2 Castiglianos Theorem on Deflections The derivation of Castiglianos theorem on deflections is based on the concept of the complimentary energy of the system The complimentary energy C is equal to the strain energy U for a linear elastic material In Castiglianos theorem, C is considered to be a function of the generalized forces (F1, F2, , Fn) acting on a system that is structurally stable (i.e. sufficiently supported)
C also depends on the distributed loads, but these loads do not need to be considered explicitly in the derivation
Castiglianos Theorem:
(5.2)
Castiglianos theorem can be extended to compute the rotation of line elements in a system subjected to couples (e.g., F1 and F2) By Eq. 5.2
(a)
(a)
Because M=bF
(5.3)
Eqs. 5.2 and 5.3 may be applied to structures with nonlinear elastic behavior
5.3 Castiglianos Theorem on Deflections for Linear Load-Deflection Relations If the structural response is assumed to be limited to linear elastic material behavior and to small displacements then the resulting load-deflection relation is linear the strain energy U equals the complimentary energy C Fig. 5.1 the Principle of Superposition applies
(5.4) (5.5)
With load-stress formulas derived for the members of the structure, U0 may be expressed in terms of the loads that act on the structure. Eqns. 5.4 and 5.5 can be used to find the displacements at the points of application of the concentrated forces (5.4) or moments
(5.5)
11
In practice, it is typically easier to find U for each type of load, then use linear superposition to combine their input to the overall U for the structure.
12
5.3.1
13
Noting that dez = ezz dz And recalling ezz = szz/E and szz = N/A, where A is the crosssectional area at z (5.8)
16
17
18
5.3.2
Origin is at left end of the beam Length of beam along the z axis x and y axes are principal axes
s zz =
Ix
19
Mx causes D*F* to rotate w.r.t. B*D* an amount df For a linear elastic material, df varies linearly w.r.t Mx The shaded area below the inclined straight line is equal to the strain energy dUM resulting from the bending of the segment dz
20
10
21
22
11
23
An exact expression for the strain energy due to the shear loading, Us, is difficult to obtain use an approximate value The shear force Vy shear stresses szy
24
12
Recall shear stress is zero on the top and bottom and maximum at the N.A. Define an average value of szy as t=Vy/A Assume this average acts over the entire beam cross section Assume the beam cross section to be rectangular and with thickness b for convenience Due to the shear force, the relative displacement of faces D*F* and B*C* is dey For linear elastic behavior, dey varies linearly w.r.t. Vy (see Fig 5.6e) dUs is the strain energy for the beam segment dz A correction (fudge) factor is introduced s.t. the expression for the shear strain energy dUs of the element is equal to kdUs
25
26
13
Exact values for k are generally not available For most problems, US is small in comparison to UM A quick approximation for k is to use ratio of the shear stress at the N.A. to the average shear stress t=Vy/A e.g., k for a rectangular cross section
(5.15)
This value is greater than the exact value (therefore is conservative as it will overpredict the situation) Authors suggest using the known value when it is available
27
To get an exact expression for Us the shear stress distribution must be known The known shear stress distribution can be used in (5.7)
28
14
5.3.3
Sections BC and DF are a distance dz apart. After the torsional loads are applied, these sections become B*C* and D*F* Section D*F* is rotated db w.r.t B*C* For linear elastic behavior, db varies linearly w.r.t. T The shaded area is the strain energy dUT
29
30
15
31
32
16
5.4
Statically Determinate
33
The strain energy for a structure is equal to the sum of the strain energies for each of its members For the jth member: Uj = UNj + UMj + USj + UTj For this chapter, loadings will be assumed to be on or about principal axes Knowing Utotal,
the deflection qi of the structure at the location of a concentrated force Fi in the direction of Fi is
(5.17)
34
17
the deflection qi of the structure at the location of a concentrated moment Mi in the direction of Mi is
(5.18)
35
Procedure for the application of Castiglianos Theorem 1. Write an expression for each of the internal actions (axial force, moment and torque) in each member of the structure in terms of the applied external loads. 2. To determine the deflection qi (or rotation qi) of the structure at the location of a concentrated force Fi (or moment Mi ) and in the directed sense of Fi (or Mi ), differentiate each of the internal action expressions w.r.t. Fi (or Mi ). 3. Substitute the expressions for internal actions obtained in Step 1 and the derivatives obtained in Step 2 into Eq. 5.17 (or Eq. 5.18) and perform the integration. The result is a relationship between the deflection qi (or rotation qi) and the externally applied loads. 4. Substitute the magnitudes of the external loads into the result obtained in Step 3 to obtain a numerical value for the displacement qi (or rotation qi).
36
18
5.4.1
The magnitude of UM for curved beams is derived in Chapter 9 The error in using Eq. 5.13 to determine UM is negligible if the radius of curvature of the beam is more than twice its depth.
37
38
19
39
40
20
41
42
21
5.4.2
So far, Castiglianos Theorem has been used to find deflections and rotations at the locations of concentrated forces and moments However, the displacement at a point other than where a concentrated load is applied may be desired
43
Procedure for the application of Castiglianos Theorem with Dummy Load 1. 2. Apply a fictitious force Fi (or moment Mi) at the point of interest for finding the displacement. Write an expression for each of the internal actions (axial force, moment and torque) in each member of the structure in terms of the applied external loadsincluding the fictitious force (or moment). To determine the deflection qi of the structure at the location of a fictitious force Fi (or Mi) and in the directed sense of Fi (or Mi), differentiate each of the internal action expressions w.r.t. Fi (or Mi). Substitute the expressions for internal actions obtained in Step 2 and the derivatives obtained in Step 3 into Eq. 5.17 (or Eq. 5.18) and perform the integration. The result is a relationship between the deflection qi (or rotation qi) and the externally applied loads, including the fictitious force Fi (or moment Mi) . Because the fictitious force (or moment) does not act on the structure, set its value to zero in the relation obtained in Step 4. Then substitute the magnitudes of the external loads into the result obtained in Step 3 to obtain a numerical value for the displacement qi (or rotation qi).
44
3.
4.
5.
22
(5.20b)
45
46
23
47
5.5
48
24
49
It can be shown that for every redundant internal force or redundant reaction (F1, F2 , )
(5.21)
To show that Eqs. 5.21 and 5.22 apply to internal forces, consider the truss with the redundant internal for NBE
50
25
Apply equal and opposite reactions NBE and NBE Because the component of the deflection of point h along member BE is not zero, it is not obvious that
(a)
51
However, NBE = NBE = NBE Likewise it can be shown for internal member forces N, V and M
(5.23)
52
26
In the application of Eqs. 5.21 and 5.22 to the system with redundant supports of redundant members, it is assumed that the unloaded system is stress-free. However, in certain applications, these conditions do not hold. Consider
Initially, the right end of the beam may be lifted off the support, or the end may exert a force on the beam due to either settling or thermal expansion (or contraction). As a result, the end of the beam (in the absence of the redundant support) may be raised a distance q1 above the location of the support before the beam is loaded or it may be a distance q2 below the support location
53
If the displacement magnitudes q1or q2 are known, then the reaction R for the loaded beam may be computed by the relations
(5.24)
Where the minus sign indicates that the displacement q1and the force R have opposite senses.
54
27
5.5.1
A structure is not altered if a redundant member or redundant support is removed and replaced by external forces and moments that are identical to the forces and moments exerted by the deleted parts. By setting the deflections at the redundant supports to known values, expressions can be determined to find the associated reactions.
55
Fig. E5.15
56
28
57
58
29
59
60
30
Fig. P5.66
61
62
31
63
64
32
65
66
33