You are on page 1of 10

.

Beam Deections by Discontinuity Functions

12

121

Lecture 12: BEAM DEFLECTIONS BY DISCONTINUITY FUNCTIONS

122

TABLE OF CONTENTS
Page

12.1. Discontinuity Functions 12.2. Examples 12.2.1. Example 1: Simply Supported Beam Under Midspan Point Load 12.2.2. Example 2: Hinged Beam Propped by Elastic Bar . . . . . 12.2.3. Example 3: SS Beam With Complicated Loading . . . . .

123 125 125 126 128

122

123 12.1. Discontinuity Functions

12.1

DISCONTINUITY FUNCTIONS

As the name indicates, Discontinuity Functions (abbreviation: DFs) were invented to compactly represent discontinuities of various kinds in mathematical functions. They may be found under several names and notations in other elds such as Physics, Chemistry and Electrical Engineering. In this course they will be used to represent beam x -functions that range from applied loads through deections. We will adopt MacAuleys angle-brackets notation, which is that used by Vable and other textbooks. The generic nomenclature is x a
n

(12.1)

where a denotes the position along x at which a discontinuity occurs. Superscript n is an integer that characterizes the kind of discontinuity represented by (12.1). (This integer can be interpreted as an exponent if nonnegative, as discussed below.) The Table provided on Figure 12.1 lists the most widely used DFs, along with their denitions.
A Short Table of Discontinuity Functions
Name doublet [Dirac] delta step ramp parabolic ramp Symbol < xa >
2 1

Definition antiderivative of doublet function is delta function antiderivative of delta function is step function 1 if x > a, else 0 xa if x > a, else 0 (xa)2 if x > a, else 0

< xa >

< xa >0 < xa >


1

2 < xa >

.............
n
th

order ramp

< xa >n

(xa)

if x > a, else 0

( n 0 )

Useful integration formula for n = 0, 1, 2, ... x < x a > n+1 < x a >n dx = valid if n 0 and x0 a n+1 x0 For beam problems where origin of x is at left end, x 0 is normally 0 If n = 1, integral is step function. If n = 2, integral is delta function.

Figure 12.1. A list of Discontinuity Functions. This Table is also provided on the Supplementary Crib Sheet for Midterm Exam #3.

It is convenient to distinguish between two types of DFs. Nonsingular DFs. Also called ordinary. For these n is nonnegative, that is, n 0. These can be graphed as conventional functions, as illustrated in Figure 12.2. As can be seen x a n can be then dened as (x a )n if x a , else 0. Consequently n can be identied as an exponent. For n = 0 the function receives several names noted in the gure. Singular DFs. If n is a negative integer, x a n is not a conventional function. [In advanced mathematics it is known as a distribution (Western literature) or generalized function (Russian 123

Lecture 12: BEAM DEFLECTIONS BY DISCONTINUITY FUNCTIONS

124

<xa>0

<xa>1 Step function, also called unit step and Heaviside function Unit ramp function unit slope a x a x

<xa>2 Parabolic unit ramp function

1 0

Figure 12.2. Discontinuity Functions for nonegative exponents. Pictured are cases n = 0, 1 and 2.

literature).] It can be dened as a limit of a sequence of functions, as illustrated in Figure 12.3. These exhibit strong singular behavior at x = a , and for that reason they are sometimes called Singularity Functions. The case n = 1 pertains to the so-called delta function or unit impulse function, which appears in many elds of engineering and sciences. Integer n is not an exponent in the usual sense but an index that identies the singularity strength. For beams a point force P (positive up) acting at x = a is mathematically representable as a scaled delta function: P x a 1 . A point moment Ma (positive CW) acting at x = a can be represented by a scaled doublet: Ma x a 2 .

<xa>

a
2

; ; ; ;
1

<xa>

Delta function, also called Dirac delta (in Physics) and unit impulse function (in dynamics)

area of hashed rectangle remains unity as 0 x a+


2

2 a

Figure 12.3. Discontinuity Functions for n = 1 (Delta function) and n = 2 (doublet). These are often called Singularity Functions in the Western literature. They do not represent conventional functions, and are dened only through a limit process as shown.

; ; ; ; ; ;

doublet function, also called dipole (in Physics) and double impulse (in dynamics) moment of hashed 2-triangle area remains unity as 0 x a +
2

For beam deection calculations, DFs are normally used in conjunction with the fourth order method, starting from the applied load p (x ). One question often asked: should reaction forces be included in p (x )? Answer: that decision is optional. If not included, they will automatically appear in the integration constants for the transverse shear Vy (x ) and the bending moment Mz (x ). The following examples indicates how the method works for several problems of varying complexity.

124

125 12.2. Examples

12.2

EXAMPLES

P
x

Constant EIzz

P A RA = P 2
(b) Support reactions

C
L/2

B RB = P 2

L/2

L (a) Problem definition

Figure 12.4. Beam problem for Example 1.

12.2.1. Example 1: Simply Supported Beam Under Midspan Point Load The problem is dened in Figure 12.4(a). Deection calculations for this conguration were done in Lecture 11 using the second order method and inter-segment continuity conditions. Here it is solved by the fourth order method and DFs, which foregoes the need for explicit continuity conditions. Write the point load as a delta function of amplitude P : p(x ) = P x 1 L 2
1

(12.2)

Integrate twice to get transverse shear and bending moment: Vy ( x ) = Mz ( x ) = p(x ) d x = P x 1 L 2 Vy ( x ) d x = P x


0

+ C1 , (12.3) C1 x + C2 .

1 L 1 2

Pause here to apply static BCs. The bending moment Mz (x ) is zero at both supports: Mz A (x ) = C2 = 0 and Mz B (x ) = P ( 1 L ) C1 L = 0 C1 = P /2. Replace into the moment expression, 2 and integrate twice more: L Mz ( x ) = P x 1 2
1

+1 P x, 2
2 3

P x1 L E Izz v (x ) = 1 2 2 E Izz v(x ) = 1 P x1 L 6 2

+1 P x 2 + C3 4 +
1 P 12

(12.4)

x 3 + C3 x + C4 ,

We now apply the two kinematic BC. The deection is zero at both supports: E Izz v A = E Izz v(0) = 1 P( 1 L )3 + 12 P L 3 + C3 L = 0 C3 = P L 2 /16. C4 = 0 and E Izz v B = E Izz v( L ) = 1 6 2 Replacing into E Izz v(x ) yields the deection curve, which can be simplied to v(x ) = P 8x1 L 2 48 E Ix x 125
3

4 x 3 + 3L 2 x

(12.5)

Lecture 12: BEAM DEFLECTIONS BY DISCONTINUITY FUNCTIONS

126

Elastic bar of D modulus E and area A Note that structure is statically determinate (visualization: if you remove the bar, it collapses) A y L x B L L w C

P
Beam of constant modulus E and inertia Izz
Figure 12.5. Structure for Example 2.

Evaluating at x = 1 L provides the midspan deection: 2 vC = v( 1 L) = 2 P L3 48 E Izz (12.6)

The procedure is faster and less error prone than that used in Lecture 11. On the other hand, it requires ability to use and manipulate DFs. 12.2.2. Example 2: Hinged Beam Propped by Elastic Bar This is a variant of the second problem discussed in the recitation of Oct 3. It is dened in Figure 12.5. Beam AC is simply supported at A and held by an elastic bar BD at half-span B. Beam AC has constant bending inertia Izz , bar BD has constant cross section A and both members have the same elastic modulus E . The beam is loaded by a point load P at C and a uniform load w over the right halfspan BC. Both P and w are taken as positive downward.
Internal bar force

F BD =2P+3wL/2
y A x B L/2 L L L/2

wL
C

RA =PwL/2

Figure 12.6. FBD of structure of Figure 12.5 showing reaction at A and internal force in bar BD.

In terms of E , A, Izz , L and P , nd: (1) axial force FB D in bar BD and reaction at A, (2) deection v B at B (this is controlled by the bar elongation) and (3) vertical deection vC at C . 126

127

12.2

EXAMPLES

To nd the axial force in bar BD, do a FBD of the structure taking moments with respect to A (to get rid of the reaction force); see Figure 12.6. This gives FB D = P + 3w L 4 (12.7)

The reaction at A can be now obtained by equating the sum of forces along y to zero: RA = P + wL 4 (12.8)

Under the internal force FB D = 2 P + (3/4)w L the bar elongates B D = FB D L (2 P + 3w L /4) L = EA EA (12.9)

This is obviouly equal to the downward beam deection at B : v B = v( L ) = B D . To nd the deection curve we start from the applied forces on the beam. for convenience we take both the reaction R A and the bar force FB D as if they were applied loads, and we keep them in compact symbolic form: p(x ) = R A x 0 Integrate twice: Vy ( x ) = Mz ( x ) = p ( x ) d x = R A + FB D x L
0 1

+ FB D x L

w x L

P x 2L

(12.10)

w x L
1

P x 2L
2

+ C1
1

V y ( x ) d x = R A x + FB D x L

1 w xL 2

P x 2L

+ C1 x + C2 .

(12.11) Note that we have replaced R A ( x 0 0 ) in the rst equation by R A x 0 = R A . This is permissible for any DF with nonnegative supercript; that is, x 0 n x n if n 0. The reason is that x cannot take negative values if the coordinate origin is placed at the left end of the beam. Pause to nd C1 and C2 . The moment at the simple support must be zero: Mz A = Mz (0) = C2 = 0 and the transverse shear must equal the negative of the reaction force: Vy A = Vy (0) = R A = R A + C1 = 0 C1 = 0. Two useful checks: MzC = M z (2 L ) = 0 and VyC = Vy (2 L ) = P . From now on the P term in (12.11) will be dropped, since for all beam points x 2 L and x 2 L vanishes if n 1. Integrating twice more: E Izz v (x ) = 1 R x2 + 1 F xL 2 A 2 BD E Izz v(x ) = 1 R x3 + 1 F xL 6 A 6 BD
2 3 n

1 w xL 6
1 w 24

+ C3
4

xL

+ C3 x + C4 .

(12.12)

The kinematic BCs are v A = v(0) = 0, because A is a simple support, and v B = v( L ) = B D = FB D L /( E A), as found above. The rst BC gives C4 = 0. The second BC, after some simplications, yields C3 = P L2 2 Izz 6 A + 127 wL 12 L2 18 Izz A . (12.13)

Lecture 12: BEAM DEFLECTIONS BY DISCONTINUITY FUNCTIONS

128

Replacing C3 , R A and FB D gives the deection curve in terms of the data as E Izz v(x ) =
1 6

P 1 wL x3 + 2 xL
4

1 6

2P +

3w L 2

xL
2

1 w 24

L2 2 Izz 6 A

wL + 12

18 Izz L A

(12.14) x.

Evaluating at x = 2 L provides the tip deection: vC = v(2 L ) = PL E 4 2L 2 + A Izz wL2 E 3 7L 2 + A 24 Izz (12.15)

Note that if A 0, that is, the bar disappears, the beam deections go to innity. 12.2.3. Example 3: SS Beam With Complicated Loading The problem is dened in Figure 12.7. The SS beam is subject to three types of applied load: (1) a uniform distributed load w over the left midspan, (2) a point force P at x = 2 L /3, and a point moment MC at midspan x = L /2. w, P and MC are positive if acting as shown. The calculation of deections will be done with the fourth order method in conjunction with DFs. Reactions are not included in the load p (x ): they will appear through the integration constants.
y

w
x

MC

P D

Constant EIzz

C
L/2 L L/6

B
L/3

Figure 12.7. Beam for Example 3

The applied load is p ( x ) = w x 0 0 + w x L = w + w x 2


0

L 2

+P x
1

2L 3

+ MC x
2

L 2

2L P x 3

L + MC x 2

(12.16)

(Note that the load term due to the point moment MC is positive if this couple acts clockwise. See Vable, p. 492, for a discussion of that topic.) Integrate twice: Vy (x ) = w x + w x Mz ( x ) = 1 w 2 x +
2 1 w 2

L 2

P x
2

2L 3

+ MC x
1

L 2

+ C1 ,
0

L x 2

2L P x 3 128

L + MC x 2

(12.17)

+ C1 x + C2 .

129

12.2

EXAMPLES

The moment at the left simple support must vanish, thus Mz A = Mz (0) = C2 = 0 and we dont need to carry C2 further. But the expression of C1 from Mz B = Mz ( L ) = 0 is involved in terms of the data, so we leave it as is for now. Integrating twice more: 2L 2 L 1 1 + MC x + 2 C1 x 2 + C3 , 3 2 L 4 1 2L 3 1 L 2 1 1 1 w x 4 + 24 w x 6P x + 2 MC x + 6 C1 x 3 + C3 x + C4 . E Izz v(x ) = 24 2 3 2 (12.18) The deection at A must be zero, which immediately gives C4 = 0. The other two integration constants are more complicated functions of the data. Setting Mz B = Mz ( L ) = 0 and then E Izz v B = E Izz v( L ) = 0 yields E Izz v (x ) = 1 w x3 + 1 w x 6 6
3

L 2

1 P x 2

C1 =

9w 8 P 24 MC , 24

C3 =

243w 512 P + 432 MC 109368

(12.19)

Substitution into E Izz v(x ) produces the deection curve


1 E Izz v(x ) = 24 w x4 +

2L 3 1 L + 2 MC x 3 2 9w 8 P 24 MC 3 243w 512 P 432 MC + x x 144 109368


1 w 24

L 2

1 P x 6

(12.20)

1 2 3 4 5 6

0.2

0.4

0.6

0.8

v(x)
(a) Deflection curve for w=100, P=0, MC =0, EIzz = 1 and L=1

0.25 0.5 0.75 1 1.25 1.5 1.75

0.2

0.4

0.6

0.8

0.75 0.5 0.25 0.25 0.5 0.2 0.4 0.6 0.8

v(x)

0.75

v(x)

(b) Deflection curve for w=0, P=100, MC =0, EIzz = 1 and L=1

(c) Deflection curve for w=0, P=0, MC =100, EIzz = 1 and L=1

Figure 12.8. Deection curves for Example 3 beam for three individual load cases.

Figure 12.8 plots deection curves for three individual load cases. Evaluating at sections C (x = 1 L ) and D (x = 2 L ) gives 2 3 1 (165 w 512 P + 240 MC ). 31104 E Izz (12.21) The midpsan deection vC is not affected by the point moment MC since this action produces an antisymmetric deection curve; see Figure 12.8(c). vC = vD = 129 1 (135 w 368 P ), 20736 E Izz

Lecture 12: BEAM DEFLECTIONS BY DISCONTINUITY FUNCTIONS

1210

The Mathematica program that solves this problem and produces the plots of Figure 12.8 is shown below. ClearAll[x,w,P,MC,C1,C2,C3,C4]; C2=C4=0; p=-w+w*UnitStep[x-1/2]-P*DiracDelta[x-2/3]; Print["p(x)=",p]; VVy=Integrate[p,x]+C1; Print["VVy=",VVy//InputForm]; Vy=-w*x+w*(x-1/2)*UnitStep[x-1/2]-P*UnitStep[x-2/3]+MC*DiracDelta[x-1/2]+C1; Mz=Integrate[Vy,x]+C2; Mz=Simplify[Mz]; Print["Mz(x)=",Mz]; EIvp=Integrate[Mz,x]+C3; EIvp=Simplify[EIvp]; Print["EIv(x)=",EIvp]; EIv=Integrate[EIvp,x]+C4;EIv=Simplify[EIv]; Print["EIv(x)=",EIv]; MzA=Mz/.x->0; MzB=Mz/.x->1; EIvA=EIv/.x->0; EIvB=EIv/.x->1; {MzA,MzB,EIvA,EIvB}=Simplify[{MzA,MzB,EIvA,EIvB}]; Print["MzA=",MzA," MzB=",MzB," EIvA=",EIvA," EIvB=",EIvB]; solC=Simplify[Solve[{MzA==0,MzB==0,EIvA==0,EIvB==0},{C1,C3}]]; Print[solC]; {C1,C3}={C1,C3}/.solC[[1]]; Print["C1=",Together[C1]," C3=",Together[C3]]; EIvx=Simplify[EIv/.solC[[1]]]; Print["EI v(x)=",EIvx]; vC=Simplify[EIvx/.x->1/2]; Print["EI vC=",Together[vC]]; vD=Simplify[EIvx/.x->2/3]; Print["EI vD=",Together[vD]]; vB=Simplify[EIvx/.x->1]; Print["EI vB=",vB]; Plot[100*EIvx/.{w->1,P->0,MC->0},{x,0,1}]; Plot[100*EIvx/.{w->0,P->1,MC->0},{x,0,1}]; Plot[100*EIvx/.{w->0,P->0,MC->1},{x,0,1}];

1210

You might also like