8 V IBRATION AND T IME R ESPONSE CHAPTER OUTLINE 8/1 8/2 8/3 8/4 8/5 8/6 Introduction Free Vibration of Particles Forced Vibration of Particles Vibration of Rigid Bodies Energy Methods Chapter Review 8/1 I N T R O D U C T I O N An important and special class of problems in dynamics concerns the linear and angular motions of bodies which oscillate or otherwise respond to applied disturbances in the presence of restoring forces. A few examples of this class of dynamics problems are the response of an engineering structure to earthquakes, the vibration of an unbalanced rotating machine, the time response of the plucked string of a musical instrument, the wind-induced vibration of power lines, and the flutter of aircraft wings. In many cases, excessive vibration levels must be reduced to accommodate material limitations or human factors. In the analysis of every engineering problem, we must represent the system under scrutiny by a physical model. We may often represent a continuous or distributed-parameter system (one in which the mass and spring elements are continuously spread over space) by a discrete or lumped-parameter model (one in which the mass and spring elements are separate and concentrated). The resulting simplified model is especially accurate when some portions of a continuous system are relatively massive in comparison with other portions. For example, the physical model of a ship propeller shaft is often assumed to be a massless but twistable rod with a disk rigidly attached to each end—one disk representing the turbine and the other representing the propeller. As a second example, we observe that the mass of springs may often be neglected in comparison with that of attached bodies. Not every system is reducible to a discrete model. For example, the transverse vibration of a diving board after the departure of the diver 569 570 Chapter 8 Vibr at i on and Ti m e Res p ons e is a somewhat difficult problem of distributed-parameter vibration. In this chapter, we will begin the study of discrete systems, limiting our discussion to those whose configurations may be described with one displacement variable. Such systems are said to possess one degree of freedom. For a more detailed study which includes the treatment of two or more degrees of freedom and continuous systems, you should consult one of the many textbooks devoted solely to the subject of vibrations. The remainder of Chapter 8 is divided into four sections: Article 8 /2 treats the free vibration of particles and Art. 8 /3 introduces the forced vibration of particles. Each of these two articles is subdivided into undamped- and damped-motion categories. In Art. 8 /4 we discuss the vibration of rigid bodies. Finally, an energy approach to the solution of vibration problems is presented in Art. 8 /5. The topic of vibrations is a direct application of the principles of kinetics as developed in Chapters 3 and 6. In particular, construction of a complete free-body diagram drawn for an arbitrary positive value of the displacement variable, followed by application of the appropriate governing equations of dynamics, will yield the equation of motion. From this equation of motion, which is a second-order ordinary differential equation, you can obtain all information of interest, such as the motion frequency, period, or the motion itself as a function of time. 8 /2 F R E E V I B R AT I O N O F P A R T I C L E S When a spring-mounted body is disturbed from its equilibrium position, its ensuing motion in the absence of any imposed external forces is termed free vibration. In every actual case of free vibration, there exists some retarding or damping force which tends to diminish the motion. Common damping forces are those due to mechanical and fluid friction. In this article we first consider the ideal case where the damping forces are small enough to be neglected. Then we treat the case where the damping is appreciable and must be accounted for. k m Equilibrium position x mg Equation of Motion for Undamped Free Vibration kx N (a) Fs Hard Linear Fs = kx Soft x (b) Figure 8/1 We begin by considering the horizontal vibration of the simple frictionless spring-mass system of Fig. 8 /1a. Note that the variable x denotes the displacement of the mass from the equilibrium position, which, for this system, is also the position of zero spring deflection. Figure 8 /1b shows a plot of the force Fs necessary to deflect the spring versus the corresponding spring deflection for three types of springs. Although nonlinear hard and soft springs are useful in some applications, we will restrict our attention to the linear spring. Such a spring exerts a restoring force −kx on the mass—that is, when the mass is displaced to the right, the spring force is to the left, and vice versa. We must be careful to distinguish between the forces of magnitude Fs which must be applied to both ends of the massless spring to cause tension or compression and the force F = −kx of equal magnitude which the spring exerts on the mass. The constant of proportionality k is called the spring constant, modulus, or stiffness and has the units N /m or lb /ft. A rt i c l e 8 / 2 F re e V i b ra t i o n o f P a r t i cl e s The equation of motion for the body of Fig. 8 /1a is obtained by first drawing its free-body diagram. Applying Newton’s second law in the form ©Fx = mẍ gives −kx = mẍ mẍ + kx = 0 or (8/1) The oscillation of a mass subjected to a linear restoring force as described by this equation is called simple harmonic motion and is characterized by acceleration which is proportional to the displacement but of opposite sign. Equation 8 /1 is normally written as ẍ + n 2x = 0 (8/2) n = √k/m (8/3) where is a convenient substitution whose physical significance will be clarified shortly. Solution for Undamped Free Vibration Because we anticipate an oscillatory motion, we look for a solution which gives x as a periodic function of time. Thus, a logical choice is x = A cos nt + B sin nt (8/4) x = C sin (nt + ) (8/5) or, alternatively, Direct substitution of these expressions into Eq. 8/2 verifies that each expression is a valid solution to the equation of motion. We determine the constants A and B, or C and , from knowledge of the initial displacement x0 and initial velocity ẋ0 of the mass. For example, if we work with the solution form of Eq. 8/4 and evaluate x and ẋ at time t = 0, we obtain x0 = A and ẋ0 = Bn Substitution of these values of A and B into Eq. 8 /4 yields x = x0 cos nt + ẋ0 sin nt n (8/6) The constants C and of Eq. 8 /5 can be determined in terms of given initial conditions in a similar manner. Evaluation of Eq. 8 /5 and its first time derivative at t = 0 gives x0 = C sin and ẋ0 = Cn cos 571 572 Chapter 8 Vibr at i on and Ti m e Res p ons e Solving for C and yields = tan−1 (x0n /ẋ0 ) C = √x02 + (ẋ0 /n ) 2 Substitution of these values into Eq. 8 /5 gives x = √x02 + (ẋ0 /n ) 2 sin [nt + tan −1 (x0n /ẋ0 )] (8/7) Equations 8/6 and 8/7 represent two different mathematical expressions for the same time-dependent motion. We observe that C = √A2 + B2 and = tan−1(A /B). Graphical Representation of Motion The motion may be represented graphically, Fig. 8 /2, where x is seen to be the projection onto a vertical axis of the rotating vector of length C. The vector rotates at the constant angular velocity n = √k/m, which is called the natural circular frequency and has the units radians per second. The number of complete cycles per unit time is the natural frequency ƒn = n /2 and is expressed in hertz (1 hertz (Hz) = 1 cycle per second). The time required for one complete motion cycle (one rotation of the reference vector) is the period of the motion and is given by = 1 /ƒn = 2 /n. x +x C ω nt ψ A B ω nt 2π τ = —– ωn C x x0 0 t 0 −C −x Figure 8/2 We also see from the figure that x is the sum of the projections onto the vertical axis of two perpendicular vectors whose magnitudes are A and B and whose vector sum C is the amplitude. Vectors A, B, and C rotate together with the constant angular velocity n. Thus, as we have already seen, C = √A2 + B2 and = tan−1(A /B). Equilibrium Position as Reference As a further note on the free undamped vibration of particles, we see that, if the system of Fig. 8 /1a is rotated 90° clockwise to obtain the system of Fig. 8 /3 where the motion is vertical rather than horizontal, 573 A rt i c l e 8 / 2 F re e V i b ra t i o n o f P a r t i cl e s the equation of motion (and therefore all system properties) is unchanged if we continue to define x as the displacement from the equilibrium position. The equilibrium position now involves a nonzero spring deflection ␦st. From the free-body diagram of Fig. 8 /3, Newton’s second law gives −k(␦st + x) + mg = mẍ k δ st k(δ st + x) x Equilibrium position m At the equilibrium position x = 0, the force sum must be zero, so that m −k␦st + mg = 0 mg Thus, we see that the pair of forces −k␦st and mg on the left side of the motion equation cancel, giving Figure 8/3 mẍ + kx = 0 which is identical to Eq. 8 /1. The lesson here is that by defining the displacement variable to be zero at equilibrium rather than at the position of zero spring deflection, we may ignore the equal and opposite forces associated with equilibrium.* k Equation of Motion for Damped Free Vibration Every mechanical system possesses some inherent degree of friction, which dissipates mechanical energy. Precise mathematical models of the dissipative friction forces are usually complex. The dashpot or viscous damper is a device intentionally added to systems for the purpose of limiting or retarding vibration. It consists of a cylinder filled with a viscous fluid and a piston with holes or other passages by which the fluid can flow from one side of the piston to the other. Simple dashpots arranged as shown schematically in Fig. 8 /4a exert a force Fd whose magnitude is proportional to the velocity of the mass, as depicted in Fig. 8 /4b. The constant of proportionality c is called the viscous damping coefficient and has units of N ∙ s/m or lb-sec /ft. The direction of the damping force as applied to the mass is opposite that of the velocity ẋ. Thus, the force on the mass is −cẋ. Complex dashpots with internal flow-rate-dependent one-way valves can produce different damping coefficients in extension and in compression; nonlinear characteristics are also possible. We will restrict our attention to the simple linear dashpot. The equation of motion for the body with damping is determined from the free-body diagram as shown in Fig. 8 /4a. Newton’s second law gives −kx − cẋ = mẍ or mẍ + cẋ + kx = 0 m c Equilibrium position x kx cx· N (a) Fd x· (8/8) (b) *For nonlinear systems, all forces, including the static forces associated with equilibrium, should be included in the analysis. mg Figure 8/4 574 Chapter 8 Vibr at i on and Ti m e Res p ons e In addition to the substitution n = √k/m, it is convenient, for reasons which will shortly become evident, to introduce the combination of constants = c/(2mn ) The quantity (zeta) is called the viscous damping factor or damping ratio and is a measure of the severity of the damping. You should verify that is nondimensional. Equation 8 /8 may now be written as ẍ + 2nẋ + n2x = 0 (8/9) Solution for Damped Free Vibration In order to solve the equation of motion, Eq. 8 /9, we assume solutions of the form x = Aet Substitution into Eq. 8 /9 yields 2 + 2n + n2 = 0 which is called the characteristic equation. Its roots are 1 = n (− + √2 − 1) 2 = n (− − √2 − 1) Linear systems have the property of superposition, which means that the general solution is the sum of the individual solutions each of which corresponds to one root of the characteristic equation. Thus, the general solution is x = A1e1 t + A2 e2 t = A1e( −+√ −1)n t + A2 e(−−√ −1)nt 2 2 (8/10) Categories of Damped Motion Because 0 ≤ ≤ ∞, the radicand (2 − 1) may be positive, negative, or even zero, giving rise to the following three categories of damped motion: I. > 1 (overdamped). The roots 1 and 2 are distinct, real, and negative numbers. The motion as given by Eq. 8/10 decays so that x approaches zero for large values of time t. There is no oscillation and therefore no period associated with the motion. II. = 1 (critically damped). The roots 1 and 2 are equal, real, and negative numbers (1 = 2 = −n). The solution to the differential equation for the special case of equal roots is given by x = (A1 + A2t)e −nt A rt i c l e 8 / 2 F re e V i b ra t i o n o f P a r t i cl e s x, mm 30 Conditions: m = 1 kg, k = 9 N/m x0 = 30 mm, x· 0 = 0 20 c = 15 N·s/m (ζ = 2.5), overdamped c = 6 N·s/m (ζ = 1), critically damped 10 0 0 1 2 3 4 t, s Figure 8/5 Again, the motion decays with x approaching zero for large time, and the motion is nonperiodic. A critically damped system, when excited with an initial velocity or displacement (or both), will approach equilibrium faster than will an overdamped system. Figure 8/5 depicts actual responses for both an overdamped and a critically damped system to an initial displacement x0 and no initial velocity (ẋ0 = 0). III. < 1 (underdamped). Noting that the radicand (2 − 1) is negative and recalling that e(a+b) = eaeb, we may rewrite Eq. 8/10 as x = 5A1 ei√1− nt + A2 e −i√1− nt 6 e−nt 2 2 where i = √−1. It is convenient to let a new variable d represent the combination n√1 − 2. Thus, x = 5A1eid t + A2 e−id t 6 e−nt Use of the Euler formula e±ix = cos x ± i sin x allows the previous equation to be written as x = 5A1 (cos dt + i sin dt) + A2 (cos dt − i sin dt)6e−nt = 5 (A1 + A2 ) cos d t + i(A1 − A2 ) sin dt6e −nt = 5A3 cos dt + A4 sin d t6 e−nt (8/11) where A3 = (A1 + A2) and A4 = i(A1 − A2). We have shown with Eqs. 8 /4 and 8 /5 that the sum of two equal-frequency harmonics, such as those in the braces of Eq. 8 /11, can be replaced by a single trigonometric function which involves a phase angle. Thus, Eq. 8 /11 can be written as x = 5C sin (d t + )6 e−nt or x = Ce−n t sin (d t + ) (8/12) 575 576 Chapter 8 Vibr at i on and Ti m e Res p ons e x, mm Conditions: m = 1 kg, k = 36 N/m c = 1 N·s /m (ζ = 0.0833) x0 = 30 mm, x· 0 = 0 – ζ ωnt Ce 30 20 Equation 8 /12 represents an exponentially decreasing harmonic function, as shown in Fig. 8 /6 for specific numerical values. The frequency d = n√1 − 2 τd 10 0 0 –10 1 3 2 4 t, s – Ce – ζ ωnt –20 is called the damped natural frequency. The damped period is given by d = 2 /d = 2/(n √1 − 2 ). It is important to note that the expressions developed for the constants C and in terms of initial conditions for the case of no damping are not valid for the case of damping. To find C and if damping is present, you must begin anew, setting the general displacement expression of Eq. 8 /12 and its first time derivative, both evaluated at time t = 0, equal to the initial displacement x0 and initial velocity ẋ0, respectively. –30 Figure 8/6 Determination of Damping by Experiment x We often need to experimentally determine the value of the damping ratio for an underdamped system. The usual reason is that the value of the viscous damping coefficient c is not otherwise well known. To determine the damping, we may excite the system by initial conditions and obtain a plot of the displacement x versus time t, such as that shown schematically in Fig. 8 /7. We then measure two successive amplitudes x1 and x2 a full cycle apart and compute their ratio 2π τd = — ωd x1 t1 x1 Ce−nt1 = = end x2 Ce−n(t1 +d) x2 t2 The logarithmic decrement ␦ is defined as t x = Ce– ζω nt sin (ωdt +ψ ) ␦ = ln a x1 2 2 b = nd = n = x2 √1 − 2 n√1 − 2 From this equation, we may solve for and obtain Figure 8/7 = ␦ √(2) 2 + ␦2 For a small damping ratio, x1 ≅ x2 and ␦ << 1, so that ≅ ␦ /2. If x1 and x2 are so close in value that experimental distinction between them is impractical, the above analysis may be modified by using two observed amplitudes which are n cycles apart. A rt i c l e 8 / 2 F re e V i b ra t i o n o f P a r t i cl e s 577 Sample Problem 8/1 A body weighing 25 lb is suspended from a spring of constant k = 160 lb /ft. At time t = 0, it has a downward velocity of 2 ft /sec as it passes through the position of static equilibrium. Determine k = 160 lb /ft (a) the static spring deflection ␦st (b) the natural frequency of the system in both rad /sec (n) and cycles /sec (ƒn) W = 25 lb (c) the system period (d) the displacement x as a function of time, where x is measured from the position of static equilibrium (e) the maximum velocity vmax attained by the mass Equilibrium position (ƒ) the maximum acceleration amax attained by the mass. Solution. mg 25 ␦st = = = 0.1562 ft or 1.875 in. k 160 n = (b) k 160 = = 14.36 rad/sec B m B 25/32.2 ƒn = (14.36)a = (c) 2 k(δ st + x) δst 1 b = 2.28 cycles/sec 2 1 1 = = 0.438 sec ƒn 2.28 ≡ mg Ans. Ans. Ans. = 0.1393 sin 14.36t ft Ans. As an exercise, let us determine x from the alternative Eq. 8 /7: = B 02 + a x0n bd ẋ0 2 (0)(14.36) 2 b sin c 14.36t + tan−1 a bd 14.36 2 = 0.1393 sin 14.36t ft (e) The velocity is ẋ = 14.36(0.1393) cos 14.36t = 2 cos 14.36t ft /sec. Because the cosine function cannot be greater than 1 or less than −1, the maximum velocity vmax is 2 ft /sec, which, in this case, is the initial velocity. Ans. (ƒ) The acceleration is ẍ = −14.36(2) sin 14.36t = −28.7 sin 14.36t ft/sec2 The maximum acceleration amax is 28.7 ft /sec2. caution in the matter of units. In the subject of vibrations, it is quite easy to commit errors due to mixing of feet and inches, cycles and radians, and other pairs which frequently enter the calculations. 2 Recall that when we refer the mo- 2 sin 14.36t 14.36 x = √x02 + (ẋ0 /n ) 2 sin c nt + tan−1 a Helpful Hints 1 You should always exercise extreme ẋ0 sin nt n = (0) cos 14.36t + mg Ans. (d) From Eq. 8 /6: x = x0 cos nt + kx x (a) From the spring relationship Fs = kx, we see that at equilibrium mg = k␦st 1 Fs = kδst Ans. tion to the position of static equilibrium, the equation of motion, and therefore its solution, for the present system is identical to that for the horizontally vibrating system. 578 Chapter 8 Vibr at i on and Ti m e Res p ons e Sample Problem 8/2 x c The 8-kg body is moved 0.2 m to the right of the equilibrium position and released from rest at time t = 0. Determine its displacement at time t = 2 s. The viscous damping coefficient c is 20 N ∙ s/m, and the spring stiffness k is 32 N/m. 8 kg Solution. We must first determine whether the system is underdamped, critically damped, or overdamped. For that purpose, we compute the damping ratio . n = √k/m = √32/8 = 2 rad/s = k Equilibrium position c 20 = = 0.625 2mn 2(8)(2) x mg Since < 1, the system is underdamped. The damped natural frequency is d = n√1 − 2 = 2√1 − (0.625) 2 = 1.561 rad /s. The motion is given by Eq. 8/12 and is cx· = 20x· x = Ce −nt sin (dt + ) = Ce −1.25t sin (1.561t + ) kx = 32x The velocity is then N ẋ = −1.25Ce−1.25t sin (1.561t + ) + 1.561Ce−1.25t cos (1.561t + ) Evaluating the displacement and velocity at time t = 0 gives x0 = C sin = 0.2 ẋ0 = −1.25C sin + 1.561C cos = 0 Helpful Hint Solving the two equations for C and yields C = 0.256 m and = 0.896 rad. Therefore, the displacement in meters is x = 0.256e −1.25t sin (1.561t + 0.896) m 1 Evaluation for time t = 2 s gives x2 = −0.01616 m. 1 We note that the exponential factor e−1.25t is 0.0821 at t = 2 s. Thus, = 0.625 represents severe damping, although the motion is still oscillatory. Ans. Sample Problem 8/3 The two fixed counterrotating pulleys are driven at the same angular speed 0. A round bar is placed off center on the pulleys as shown. Determine the natural frequency of the resulting bar motion. The coefficient of kinetic friction between the bar and pulleys is k. Solution. The free-body diagram of the bar is constructed for an arbitrary ω0 Central position a –– 2 y A G displacement x from the central position as shown. The governing equations are [©Fx = mẍ ] k NA − k NB = mẍ [©Fy = 0] NA + NB − mg = 0 1 [©MA = 0] aNB − a μk NA NA a + xb mg = 0 2 ω0 a x a –– 2 mg B μk NB NB Helpful Hints Eliminating NA and NB from the first equation yields 2 1 Because the bar is slender and does 2k g ẍ + x=0 a not rotate, the use of a moment equilibrium equation is justified. We recognize the form of this equation as that of Eq. 8 /2, so that the natural frequency in radians per second is n = √2k g/a and the natural frequency in cycles per second is 2 We note that the angular speed 0 ƒn = 1 √2k g/a 2 Ans. does not enter the equation of motion. The reason for this is our assumption that the kinetic friction force does not depend on the relative velocity at the contacting surface. A r ti c l e 8 / 2 P r o b l e ms PROBLEMS (Unless otherwise indicated, all motion variables are referred to the equilibrium position.) 579 8/5 For the spring-mass system shown, determine the static deflection ␦st, the system period , and the maximum velocity vmax which result if the cylinder is displaced 100 mm downward from its equilibrium position and released from rest. Undamped, Free Vibrations 8/1 When a 3-kg collar is placed upon the pan which is attached to the spring of unknown constant, the additional static deflection of the pan is observed to be 42 mm. Determine the spring constant k in N /m, lb /in., and lb /ft. Problem 8/5 8/6 The cylinder of the system of Prob. 8 /5 is displaced 100 mm downward from its equilibrium position and released at time t = 0. Determine the position y, velocity v, and acceleration a when t = 3 s. What is the maximum acceleration? Problem 8/1 8/2 Determine the natural frequency of the spring-mass system in both radians per second and cycles per second (Hz). 8/7 Determine the natural frequency in cycles per second for the system shown. Neglect the mass and friction of the pulleys. Assume that the block of mass m remains horizontal. x k k = 54 lb/in. 64.4 lb Problem 8/2 8/3 For the system of Prob. 8 /2, determine the position x of the mass as a function of time if the mass is released from rest at time t = 0 from a position 2 inches to the left of the equilibrium position. Determine the maximum velocity and maximum acceleration of the mass over one cycle of motion. 8/4 For the system of Prob. 8 /2, determine the position x as a function of time if the mass is released at time t = 0 from a position 2 inches to the right of the equilibrium position with an initial velocity of 9 in. /sec to the left. Determine the amplitude C and period of the motion. m Problem 8/7 580 Chapter 8 Vibr at i on and Ti m e Res p ons e 8/8 The vertical plunger has a mass of 2.5 kg and is supported by the two springs, which are always in compression. Calculate the natural frequency ƒn of vibration of the plunger if it is deflected from the equilibrium position and released from rest. Friction in the guide is negligible. k1 = 3.6 kN/m Problem 8/10 Fixed k2 = 1.8 kN/m 8/11 For the cylinder of Prob. 8 /10, determine the vertical displacement x, measured positive down in millimeters from the equilibrium position, in terms of the time t in seconds measured from the instant of release from the position of 25 mm added deflection. 2.5 kg Problem 8/8 8/12 Determine the natural frequency in radians per second for the system shown. Neglect the mass and friction of the pulleys. 8/9 Determine the period for the system shown. The cable is always taut, and the mass and friction of the pulley are to be neglected. k k m m k Problem 8/9 8/10 In the equilibrium position, the 30-kg cylinder causes a static deflection of 50 mm in the coiled spring. If the cylinder is depressed an additional 25 mm and released from rest, calculate the resulting natural frequency ƒn of vertical vibration of the cylinder in cycles per second (Hz). Problem 8/12 8/13 An old car being moved by a magnetic crane pickup is dropped from a short distance above the ground. Neglect any damping effects of its worn-out shock absorbers and calculate the natural frequency ƒn in cycles per second (Hz) of the vertical vibration which occurs after impact with the ground. Each of the four springs on the 1000-kg car has a constant of 17.5 kN/m. Because the center of mass is located midway between the axles and the car is level when dropped, there is no rotational motion. State any assumptions.
0
You can add this document to your study collection(s)
Sign in Available only to authorized usersYou can add this document to your saved list
Sign in Available only to authorized users(For complaints, use another form )