Jtam.dvi JOURNAL OF THEORETICAL AND APPLIED MECHANICS 46, 1, pp. 51-68, Warsaw 2008 ON THE ANALYTICAL SOLUTION FOR THE SLEWING FLEXIBLE BEAM-LIKE SYSTEM: ANALYSIS OF THE LINEAR PART OF THE PERTURBED PROBLEM André Fenili National Institute for Space Research (INPE), São José dos Campos, Brazil University of Taubaté, Department of Engineering Mechanics, Taubaté, SP, Brazil e-mail: fenili@dem.inpe.br; fenili@unitau.br José Manoel Balthazar Departamento de Estat́ıstica Matemática Aplicada e Computação, UNESP, Rio Claro, SP, Brazil Mechanical Design Department, UNICAMP, Campinas, SP, Brazil e-mail: jmbaltha@rc.unesp.br The dynamical system investigated in this work is a nonlinear flexible beam-like structure in slewing motion. Non-dimensional and perturbed governing equations ofmotion are presented. The analytical solution for the linear part of these perturbed equations for ideal and for non-ideal cases are obtained. This solution is necessary for the investigation of the complete weak nonlinear problemwhere all nonlinearities are small perturbations around a linear known solution. This investigation shall help the analyst in the modelling of dynamical systems with structure- actuator interactions. Key words: non-ideal system, flexible structure, perturbed equations, analytical solution 1. Introduction The study of dynamic behaviour (and control) of slewing flexible structures has in view the improvement of lightweight and faster structures. These in- vestigations are complex and present continuing interest from researchers and scientists. 52 A. Fenili, J.M. Balthazar The applications of the theory of slewing flexible structures may be divi- ded basically into two groups: robotics and aerospace structure applications. Typically, themodal analysis approach is themost popularmodel approach to aerospace slewing structures, and the finite element approach is the onemost frequently used to investigate robotic manipulators. Low inherent damping, small natural frequencies, and extreme light weights are some common cha- racteristics of these systems, and which make them vulnerable to any exter- nal/internal disturbances (such as angular maneuvers, impacts, etc). Robot arms with such characteristics are easy to carry out, need smaller actuators and can reach objectives in a greaterworkspace since they are thinner and lon- ger than the rigid ones usually used for the same task. DCmotors are popular actuators for lightweight manipulators not only because they can generate a wide range of torque and angular velocity, but also because they are quiet, clean and efficient. Nowadays, in the competitive world, the search for such kind of mechanical systems is an increasing preoccupation. Many of the publishedpapers in this area are concernedwith the dynamics and/or control of flexible beam-like structures. Little effort has been focused on the actuator-structure interaction. This interaction affects the dynamics of the whole slewing system (actuator included). The goal of this work is to deal with this interaction. In systems such as lightweight robotic manipulators, solar panels and antennas in satellites, helicopter blades and so on, the mutual interaction between the angular di- splacement of the slewing axis and the flexible structure deflection can be very important inhighangular speedmaneuvers (Balthazar et al., 1999, 2001-2004). Dynamical systems, in which this interaction occurs, are called non-ideal systems (Kononenko, 1969). In the ideal system, there is nomutual interaction and only the actuator dynamics excites the structure dynamics. Governing equations of motion for the ideal and non-ideal damped beam- like slewing flexible structure connected to a DC motor are presented with discussions on nonlinear effects in Fenili (2004), Fenili et al. (2004). These equations and boundary conditions are nondimensional and scaled quantities also known as perturbed equations of motion because of the small parameter which multiplies nonlinear and damping terms (Fenili, 2000; Fenili and Bal- thazar, 2005a,b). Some experimental results and discussions were related in Fenili et al. (2001). Next, some papers are mentioned from the current literature concerning the subject treated here, in order to understand better the position of this problem in themain current literature. On the analytical solution for the slewing... 53 2. A brief literature review Slewing flexible structures were first considered, in literature, by Book et al. (1975). In that paper, the authors applied amodal truncatedmodel on a sys- tem composed by two flexible beams with two joints, and included discussion on the controller design with a torque source. Considerable efforts have been performed since then by other authors in this direction. Some of those authors arementioned next (without undeservingmany others notmentioned) in this paper. Cannon and Schmidt (1984) developed and demonstrated stable and pre- cise position control of one end of a very flexible beam by using direct me- asurements from the free end position as a basis for torquing at the other end. The authors demonstrated that a satisfactory feedback tip-control re- sponse can be achieved with a good dynamic model of a flexible arm. Bayo (1987) analyzed a structural finite element algorithm on the linear Bernoulli- Euler beam in order to calculate the end torque necessary to produce desired motion at the free tip of a flexible link. The computed torque, smaller than the one required for the rigid link of some weight, provided desired tip mo- tion without overshoot. The authors used this approach in both open loop and feedback controls. Juang et al. (1986) studied a slewing flexible structu- re experimentally; Juang and Horta (1987) developed a hardware set-up to study slewing control for slewing flexible structures and using linear optimal control to design active control (implemented in an analog computer); Juang et al. (1989) discussed several important issues related to slewing experiments with flexible structures including nonlinearities and calibration of actuators and sensors. In Hamilton et al. (1991) a simple experimental set-up wasmade up by a pair of single-axis flexible beams attached to a DC servo motor in order to illustrate collocated and noncollated control for this kind of struc- tures. An optical encoder and strain gauges provided hub and beam position information, respectively. Their results on non-collocated zero-placement con- trol illustrate that the feedback from the beam and motor hub provides the necessary information for vibration suppression in slewing maneuvers. Yang et al.) (1994) realized numerical simulations concerning slewing control tasks for a planar articulated double-beam structure. Garcia and Inman (1990) in- vestigated the modeling of a single link flexible slewing beam torque driven by a DC motor at the slewing axis. Effects of modal participation factors in the slewing equations of motion were discussed. These factors were indicative of the degree into which actuators and flexible structures interacted with one another dynamically. Boundary conditions of the slewing beamwere determi- 54 A. Fenili, J.M. Balthazar ned by the actuator that drove the entire system. Garcia and Inman (1991) considered the slewing control of an active flexible structure by examining the governing equations of motion of an integrated actuator-structure system composed by a thin aluminum beam torque driven by an armature controlled electric motor and actuated by a piecewise distributed piezoceramic actuator. Their approach offered an advantage of reducing the peak voltage demands on the motor. In addition, that active structure approach substantially redu- ced the maximum tip deflection of the beam. Sah et al. (1993), using a finite element algorithm, considered the closed-loop performance and the dynamic interaction between a DC motor and a slewing beam. The authors conclu- ded that systems with an appropriate amount of actuator-beam interaction tend to be more easily controlled and require a modest amount of actuator efforts, but systems with little actuator-beam interaction are especially prone to beam vibrations and require feedback of beam dynamics for good closed- loop performance. It was also shown that systems with excessive interaction require stabilization efforts in order to obtain good transient performance and tend to consume increased levels of actuation energy. In Kwak et al. (1994), a slewing flexible beam equippedwith piezoelectric sensors and actuators was modeled via extended Hamilton’s principle considering nonlinearities coming from rigid body rotational motion, whereas the elastic vibration was assumed to be small compared to rigid body rotation. The authors used a decentralized control technique in which the control of the whole system was divided into the slewing control and the vibration suppression control. The sliding model control was proposed as the slewing control, and themodal space positive fe- edback plus disturbance-counteracting control was developed for suppression of vibrations. These techniques were verified by experiments. The experimen- tal results showed that the decentralized control performs satisfactorily, but high frequency vibrations remained uncontrolled. In this case, there will be a need for more actuators, sensors andmore powerful piezoelectric actuators. Finally, Fenili (2000) has started the studyof nonlinear dynamicbehaviour of flexible structures in planar slewingmotions for linear and nonlinear curva- ture assumptions. A schematic of the system investigated there (and here) is depicted in Fig.1. In this paper, the authors present a search for an approxi- mate analytical solution to this problem. A short version of this subject was discussed in Fenili and Balthazar (2005a). An experimental apparatus was constructed in order to verify the theoretical developments and is shown in Fig.2 (Fenili et al., 2001). This paper is organized as follows. In Section 3 the governing equations of motion are presented. In Section 4 the analytical solution of the linear non- On the analytical solution for the slewing... 55 Fig. 1. Schematic of a slewing beam-like flexible structure (Fenili, 2000) Fig. 2. Experimental set-up of the slewing flexible beam-like structure (Fenili, 2000; Fenili et al., 2001) ideal problem(onemodeexpansion) isdeveloped. InSection5 someconcluding remarks are presented. In Section 6 someacknowledgements aremade. Finally, main bibliographic references used in this work and two appendixes in order to detail and/or give values for the expressions, equations and coefficients presented along the paper are listed. 3. Governing equations of motion Thegoverning equations ofmotion for the systemdepicted inFig.1 are derived from the extended Hamilton principle (Fenili, 2000; Fenili et al., 2004a). This mathematical model is based on a nonlinear curvature assumption for the beam. The nonlinear equations obtained are put on a dimensionless form and 56 A. Fenili, J.M. Balthazar scaled in order to become perturbed equations of the kind (Nayfeh andMook, 1973) linear terms + ǫ2(nonlinear terms + damping) = 0 (3.1) The small parameter, ǫ, in (3.1) is given by ǫ= r2 L2 (3.2) where r is the radius of gyration of the cross-section area of the beamand L is the length of the beam.The perturbed equations are then discretized through the assumed method of modes. On this assumption, the deflection variable, v(x,t), is written as v(x,t) = n ∑ i=1 φi(x)qi(t) (3.3) where n is the number of modes, φi(x) is any admissible function which represents the assumed solutions in space and qi(t) are unknown solutions in time. Considering the longitudinal deflection of the beam, u(x,t), as being of the order O(ǫ2), assuming θ as the slewing angle and ia as the armature current in the DC motor, the perturbed discretized governing equations of motion for thenon-ideal system,after somemanipulations inorder to eliminate u(x,t) from the set of governing equations (Fenili, 2000), are given by i̇a+ c1ia+ c2θ̇= c1U θ̈+ c3θ̇− c4ia− c5φ ′′ 1(0)q1 =0 (3.4) q̈1+w 2 1q1+α1θ̈+ ǫ 2[µq̇1+β11θ̇ 2q1−℘111θ̇q1q̇1+λ111θ̈q 2 1 + +Λ1111q1q̇ 2 1 +Λ1111q 2 1q̈1+Γ1111q 3 1] = 0 The boundary conditions are given by φ(0, t) = 0 φ′(0, t) = 0 φ′′(1, t) = 0 φ′′′(1, t) = 0 The coefficients of Eqs. (3.4) are presented inAppendixA. For the linear ideal problem, one must consider c5 =0 in Eq. (3.4)2. 4. Analytical solution to the linear problem (one mode expansion) For the linear system (ǫ = 0 in Eq. (3.4)3) associated with the primary re- sonance of the first flexural mode, one needs to make U = 0 in Eq. (3.4)1. On the analytical solution for the slewing... 57 To run numerical simulations, the initial condition is given in the variable q1 (associated to the time behaviour of the beam deflection), as can be seen in Figures 3 to 5. Considering one mode expansion (n=1), the set of equations to be ana- lytically solved is given by i̇a+ c1ia+ c2θ̇=0 θ̈+ c3θ̇− c4ia− c5φ ′′ 1(0)q1 =0 (4.1) q̈1+w 2 1q1+α1θ̈=0 Themode shapes considered, are given by φi(x)= cosh(aix)− cos(aix)−αi[sinh(aix)− sin(aix)] where αi = cosh(aiL)+cos(aiL) sinh(aiL)+sin(aiL) and ai is associated with the eigenvalues of the clamped-free undamped and not excited linear (Bernoulli-Euler) beam. 4.1. Analytical solution for q1(t) Dividing Eq. (4.1)3 by −α1, adding to Eq. (4.1)2 and solving for ia, gives ia =− ( 1 α1c4 ) q̈1− ( w21 α1c4 + c5φ ′′ i (0) c4 ) q1+ (c3 c4 ) θ̇ (4.2) Differentiating (4.2) with respect to time, yields i̇a =− ( 1 α1c4 )... q 1− ( w21 α1c4 + c5φ ′′ i (0) c4 ) q̇1+ (c3 c4 ) θ̈ (4.3) Substituting Eqs. (4.2) and (4.3) into (4.1)1, results − ( 1 α1c4 )... q 1− ( c1 α1c4 ) q̈1− ( w21 α1c4 + c5φ ′′ i (0) c4 ) q̇1− (c1w 2 1 α1c4 + c1c5φ ′′ i (0) c4 ) q1+ (4.4) + (c3 c4 ) θ̈+ (c1c3 c4 +c2 ) θ̇=0 Solving Eq. (4.1)3 for θ̈, gives θ̈=− ( 1 α1 ) q̈1− (w21 α1 ) q1 (4.5) 58 A. Fenili, J.M. Balthazar Integrating Eq. (4.5) from 0 to t, are obtains: θ̇=− ( 1 α1 ) q̇1− (w21 α1 ) t ∫ 0 q1 dt (4.6) SubstitutingEqs. (4.1)1,2 into (4.4), differentiating the resulting equationwith respect to time andmultiplying by −1, results A .... q 1+B1+Cq̈1+Dq̇1+Eq1 =0 (4.7) where: A= 1 α1c4 B= c1+c3 α1c4 C = w21 + c1c3 α1c4 + c5φ ′′ 1(0) c4 + c2 α1 D= (c1+ c3)w 2 1 α1c4 + c1c5φ ′′ 1(0) c4 E = c1c3w 2 1 α1c4 + c2w 2 1 α1 Equation (4.7) is a fourth order ordinary differential equation with constant coefficients in the variable q1. The characteristic equation associated to (4.7) is given by Ar4+Br3+Cr2+Dr+E=0 (4.8) If Eq. (4.8) has only real and not equal roots, a solution to (4.7) is of the form q1 =C1e r1at+C2e r2at+C3e r3at+C4e r4at (4.9) If Eq. (4.8) has only complex roots, the solution to (4.7) is of the form q1 =C1e r1atcos(r1bt)+C2e r2at sin(r2bt)+C3e r3atcos(r3bt)+C4e r4at sin(r4bt) (4.10) If Eq. (4.8) has two complex and two real roots, the solution to (4.7) is of the form q1 =C1e r1atcos(r1bt)+C2e r2at sin(r2bt)+C3e r3at+C4e r4at (4.11) where ria and rib are associated which each of the possible complex roots of (4.8). Let the initial time be represented by tr. Equations (4.1) can be rewritten in tr, resulting i̇ar =−c1iar − c2θ̇r θ̈r =−c3θ̇r+ c4iar + c5φ ′′ i (0)q1r (4.12) q̈1r =−w 2 1q1r−α1θ̈r On the analytical solution for the slewing... 59 Differentiating (4.12)2,3 with respect to time, one finds ... θ r and ... q 1r. Using Eqs. (4.9), (4.10) or (4.11) and their time derivatives (in t= tr), a set of algebraic equations can be formulated. From these equations, one can find the constants Ci. With the values in Appendix B, one can easily find A, B, C, D and E, and calculate roots of the characteristic equations. The solution given by Eq. (4.11), for each case, is given by q1case1 =1.1958 ·10 4e−24.7510t cos(962.4100t)+345.9672e−24.7510t · ·sin(962.4100t)+60.3107e−613.4746t +2.0655e−6.4628·10 −5t q1case2 =−11.4037e −47.1299t cos(78.5155t)+4.9359e−47.1299t sin(78.5155t)− −0.0433e−592.6730t +8.4131 ·104e−0.0113t (4.13) q1case3 =7.8895 ·10 5e−0.5874t cos(0.8169t)−6.3616 ·105e−0.5874t sin(0.8169t)+ +15.2815e−591.4816t −0.1039 ·105e−95.4528t Fig. 3. Non-dimensional transverse deflection of the versus time – case 1 Fig. 4. Non-dimensional transverse deflection of the versus time – case 2 Figures 3 to 5 illustrate solutions (4.13). These solutions are stable and damped. 60 A. Fenili, J.M. Balthazar Fig. 5. Non-dimensional transverse deflection of the versus time – case 3 5. Analytical solution for θ(t) Integrating Eq. (4.6) in time and using solution (4.11), are obtains θ=E1e r1atcos(r1bt)+E2e r1at sin(r1bt)+E3e r2atcos(r2bt)+ (5.1) +E4e r2at sin(r2bt)+E5e r3at+E6e r4at+E7t+E8 where E0 = C1w 2 1(r 2 1a−r 2 1b) α1(r 2 1a+r 2 1b )2 E1 =− C1+E0α1 α1 E2 =− 2C1w 2 1r1ar1b α1(r 2 1a+r 2 1b )2 E3 =− 2C2w 2 1r2ar2b α1(r 2 2a+r 2 2b )2 E4 =− C2((r 2 2a+r 2 2b) 2+w21(r 2 2a−r 2 2b)) α1(r 2 2a+r 2 2b )2 E5 =− C3(r 2 3a+w 2 1) α1r 2 3a E6 =− C4(r 2 4a+w 2 1) α1r 2 4a E7 = w21 α1 ( C1r1a r21a+r 2 1b + C3 r3a + C4 r4a − C2r2b r22a+r 2 2b ) E8 = w21 α1 (C3 r23a + C4 r24a + E3α1 w21 ) +E0 Figures 6 to 8 show solution (4.15). For case 1, the angular displacement is damped. For cases 2 and 3, the oscillation of the beam is damped but the angular displacement increases in time. On the analytical solution for the slewing... 61 Fig. 6. Angular displacement (nondimensional) – case 1 Fig. 7. Angular displacement (nondimensional) – case 2 Fig. 8. Angular displacement (nondimensional) – case 3 5.1. Analytical solution for i a (t) Substituting q1 (Eq. (4.11)), its second timederivative and θ̇ (usingEq. (4.2)), gives ia =(H1cos(r1bt)+H2 sin(r1bt))e r1at+[H3 sin(r2bt)+H4cos(r2b)]e r2at+ (5.2) +H5e r3at+H6e r4at+H7+H8 62 A. Fenili, J.M. Balthazar where H1 =− C1(r 2 1a−r 2 1b+w 2 1) α1c4 − C1c50+ c3A(E1r1a+E2r1b) c4 c50 = c5φ ′′ i (0) H2 = 2C1r1ar1b α1c4 −H2b H2b = c3A(E1r1b−E2r1a) c4 H3 =− C2(r 2 2a−r 2 2b+w 2 1) α1c4 − C2c50− c3B(E3r2b−E4r2a) c4 H4 =− 2C2r2ar2b α1c4 + c3B(E3r2a+E4r2b) c4 H5 =H5b− C3c50+ c3E5r3a c4 H5b =− C3(r 2 3a+w 2 1) α1c4 H6 =− C4(r 2 4a+w 2 1) α1c4 − C4c50+ c3E6r4a c4 H7 = c3E7 c4 H8 =−H1−H4−H5−H6−H7 Fig. 9. Armature current (nondimensional) – case 1 The constant H8 is introduced in order to guarantee that ia(0) = 0. Fi- gures 9 to 11 illustrate the solution given by Eq. (4.16) to each of the studied cases. On the analytical solution for the slewing... 63 Fig. 10. Armature current (nondimensional) – case 2 Fig. 11. Armature current (nondimensional) – case 3 6. Conclusions The analytical solution to the set of differential equations associated with the linear part of the perturbed governing equations of motion for the general problem under investigation (nonlinear slewing flexible beam-like structures) is presented here. This solution, considered of the order ǫ0, is very important to start the search for a perturbed solution to the weak nonlinear problem presented inEqs. (3.4). The analytical solutions obtainedhere for the actuator and for the beam (and plotted in Fig. 3 to 11) are stable. The next step in this investigation consists in applying a perturbation technique to solve the perturbed problem using the solution presented here (a remarkable task!). Using themultiple scalemethod, for example, the nonlinear perturbed equations of motion of the order ǫ2 to be solved now are given by ∂ia0 ∂T0 + c1ia0+ c2 ∂θ0 ∂T0 + ǫ2 [∂ia1 ∂T0 + ∂ia0 ∂T1 + c1ia1+c2 ∂θ1 ∂T0 + c2 ∂θ0 ∂T1 ] = = ǫ2 [uc1 2i (eiΩt− e−iΩt) ] 64 A. Fenili, J.M. Balthazar ∂2θ0 ∂T20 + c3 ∂θ0 ∂T0 − c4ia0− c5φ ′′ 1(0)q10+ +ǫ2 [∂2θ1 ∂T20 + c3 ∂θ1 ∂T0 − c4ia1− c5φ ′′ 1(0)q11+ c3 ∂θ0 ∂T1 +2 ∂θ0 ∂T0∂T1 ] =0 ∂2q10 ∂T20 +w21q10+α1 ∂2θ0 ∂T20 + ǫ2 [∂2q11 ∂T20 +w21q11+α1 ∂2θ1 ∂T20 +β11q10 (∂θ0 ∂T0 )2 + −℘111q10 ∂θ0 ∂T0 ∂q10 ∂T0 +λ111q 2 10 ∂2θ0 ∂T20 +Λ1111q10 (∂q10 ∂T0 )2 + +Λ1111q 2 10 ∂2q10 ∂T20 +Γ1111q 3 10+2α1 ∂θ0 ∂T0∂T1 +2 ∂q10 ∂T0∂T1 ] =0 A. The coefficients of equations (3.4) c1 = RaT La c2 =Ng c3 = cvN 2 gT It c4 = NgKtKbT 2 LmIt c5 = EIT2 LIt It = Ishaft +N 2 gImotor Rij = x ∫ 0 φ′i(ξ)φ ′ j(ξ) dξ=Rji Vi =− 1 ∫ x φi(ξ) dξ Λijkℓ = 1 ∫ 0 (Sjkφ ′′ iφℓ+Rjkφ ′ iφℓ) dx αℓ = 1 ∫ 0 xφℓ dx Sij =− 1 ∫ x [ η ∫ 0 φ′i(ξ)φ ′ j(ξ) dξ ] dη Wij =− 1 ∫ x φ′i(ξ)φj(ξ) dξ ℘ijℓ =2 1 ∫ 0 (Rijφℓ−φ ′′ iVjφℓ−φ ′ iφjφℓ) dx βiℓ = [ 1 ∫ 0 ( xφ′iφℓ+ 1 2 (x2−1)φ′′iφℓ ) dx ] −1 On the analytical solution for the slewing... 65 λijℓ = 1 ∫ 0 ( − 1 2 Rijφℓ+φ ′′ iVjφℓ+φ ′ iφjφℓ ) dx Γijkℓ = 1 ∫ 0 [ 3 1.87804 φ′iφ ′′ jφ ′′′ kφℓ+ 3 2(1.87804) φ′′iφ ′′ jφ ′′φℓ+ +w2j(φ ′ iφjφ ′ kφℓ+Wijφ ′′ kφℓ) ] dx where: Ra represents the armature resistance, T is the period of the first natural frequency of the beam, La – armature inductance, cv –motor internal damping, Ishaft – inertia of the connectingmotor-beam shaft, Imotor – inertia of the motor, Kt – torque constant, Kb – back e.m.f. constant, E – Young’s modulus, I – inertia of the beam cross section around the neutral axis, L – beam length, φℓ – each of the flexural vibrationmode and wℓ – corresponding model frequency. B. Beam properties (cases 1 to 3) and motor parameters T1 =1 w1 =1 Ieixo =0.0000369kgm 2 α1 =0.570157 φ ′′ i (0)= 7.05377 B.1. Beam Case 1: aluminum, L= 0.20m, E = 0.70 ·1011N/m2, height = 0.02544m, height = 0.02544m, A = 0.00008039m2, I = 6.6896 · 10−11m4, ρ=2700kg/m3, ǫ=0.000020804 Case 2: aluminum, L= 1.40m, E = 0.70 ·1011N/m2, height = 0.02544m, basis = 0.00316m, A = 0.00008039m2, I = 6.6896 · 10−11m4, ρ=2700kg/m3, ǫ=0.00000042456 Case 3: steel, L = 0.8720m, E = 2.10 · 1011N/m2, height = 0.01587m, basis = 0.00082m, A = 0.00001301m2, I = 7.2918 · 10−13m4, ρ=7800kg/m3, ǫ=0.000000073710 Obs.: ”height” and ”basis” refer to the beam cross section (Fenili,2000). 66 A. Fenili, J.M. Balthazar B.2. DC motor parameters Cm = 0.0046290Nms/rad, Kt = 0.0528140Nm/A, Ra = 1.9149520Ω, Kb =0.0528140Vs/rad, Lm =0.0031000H, Imotor =0.0000654kgm 2. Acknowledgements The authors wish to thank Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) and Conselho Nacional de Pesquisas (CNPq) for the grants that supported this study. References 1. Balthazar J.M., Mook D.T., Brasil R.M.L.R.F., Weber H.I., Fenili A.,BelatoD.,Felix J.L.P., 2001,Recent results onvibratingproblemswith limited power supply, In: Sixth Conference on Dynamical Systems Theory and Applications, Awrejcewicz J.,Grabski J., Nowakowski J. (Edit.), Lodz, Poland, 27-50 2. Balthazar J.M., Mook D.T., Weber H.I, Brasil R.M.L.R.F., Fenili A., Belato D., Felix J.L.P., 2003, An overview on non-ideal vibrations, Meccanica, 38, 613-621 3. Balthazar J.M., Mook D.T., Weber H.I, Brasil R.M.L.R.F., Fenili A., Belato D., Felix J.L.P.,Garzeri F., 2004,A review on new vibration issues due to non-ideal energy sources, In: Dynamics Systems and Control, Udwadia F.E., Weber H.I., Leitman G. (Edit.), Stability and Control: Theory, Method and Applicatuions, 22, Chapman&Hallick, 237-258 4. Balthazar J.M., Mook D.T., Weber H.I., Fenili A., Belato D., de Mattos M.C., Wieczorek S., 1999, On vibrating systems with a limited power supply and their applications to Engineering sciences, Short Course pre- sented in Seminário Brasileiro de Análise, UNICAMP, Campinas, São Paulo, Brasil, May, 137-227 5. Bayo E., 1987, A finite approach to control the end-point of a single link flexible robot, J. of Robotic Systems, 4, 1, 63-67 6. Book W.J., Maizza-Neto O., Whitney D.E., 1975, Feedback control of two beam, two joint systems with distributed flexibility, J. Of Dynamic Sys- tems, Measurement and Control, 429-431 7. CannonR.H., SchmidtE., 1984, Initial experimentson the end-point control of a flexible one-link robot, International Journal of Robotics Research, 3, 3, 62-75 On the analytical solution for the slewing... 67 8. FeniliA., 2000,MathematicalModeling andAnalysis of the Ideal andNonideal Behavior of Slewing Flexible Structures, Ph.D. Thesis, University of Campinas, Department of Engineering, Brazil [in Portuguese] 9. Fenili A., 2004, Modeling, identification and control of linear and nonline- ar slewing flexible structures with aerospace applications (satellites and robo- tics), Volume 1 andVolume 2,TechnicalReport, (PRP) INPE-11829-PRP/244 and (PRP) INPE-11830-PRP/245, Instituto Nacional de Pesquisas Espaciais (INPE), São José dos Campos [in Portuguese] 10. Fenili A., Balthazar J.M., 2005a, On the analytical solution of the linear problem associated to the perturbed equations of a nonlinear slewing flexible structure: ideal and nonideal, 8th Conference on Dynamical Systems Theory and Applications, Łódź, Poland 11. Fenili A., Balthazar J.M., 2005b, Some remarks on nonlinear vibrations of ideal and non-ideal slewing flexible structures, Journal of Sound andVibration, 268, 800-841 12. Fenili A., Balthazar J.M., Brasil R.M.L.R.F., 2004, On the mathe- matical modeling of beam-like flexible structure in slewing motion assuming nonlinear curvature, Journal of Sound and Vibrations, 282, 543-552 13. Fenili A., Balthazar J.M., Mook D.T., 2001, Some remarks about the experimental analysis of slewing flexible structure and mathematical mode- ling, Proceedings of the IX DINAME, Florianopolis, Edited by J.J. Espindola, E.M.O. Lopes, F.S.C. Bazan, 41-346 14. Garcia E., InmanD.J., 1990,Modeling actuator-structure interaction in the slewing of flexible structures,Proceedings of theAmericanControls Conference, San Diego, California, 1962-1967 15. Garcia E., Inman D.J., 1991, Advantages of slewing an active structure, J. of Intell. Mater. Syst. and Struct., 261-272 16. Hamilton J., Lee K.F.G., Juang J.N., 1991, A single testbed for slewing control experiments,American Controls Conference, 1733-1737 17. Juang J.N., Horta L.G., 1987, Effects of atmosphere on slewing control of a flexible structure, J. Guidance, 1094, 387-392 18. Juang J.N., Horta L.G., RobertshawH.H., 1986,A slewing control expe- riments for flexible structures, Journal of Guidance, Control and Dynamics, 9, 599-607 19. Juang J.N., Yang L.-F., Huang J.K.,MacauleyR., 1989,Rapid rotatio- nal/translational maneuvering experiments of a flexible steel beam, American Controls Conference, 1403-1408 20. Kononenko V.O., 1969,Vibrating Systems with Limited Power Supply, Illife, London 68 A. Fenili, J.M. Balthazar 21. Kwak M.K., Denoyer K.K., Sciulli D., 1994, Dynamics and control of slewing active beam, J. Guidance, 18, 1, 185-189 22. Nayfeh A.H., Mook D.T., 1973, —em Nonlinear Oscillations, Wiley, New York 23. Sah J.J.F., Lin J.S., Crassidis A.L., Mayne R.W., 1993, Dynamic inte- raction in a slewing motor-beam system: closed loop control, ASME PAPER DE, 61, 257-263 24. Yang L.-F., Chew M.-S., Juang J.N., 1994, Concurrent mechanism and control design for the slewing of flexible space structures,Journal ofMechanical Design, 116, 944-951 O analitycznym rozwiązaniu dla układu podatnej belki obrotowej: analiza części liniowej zagadnienia perturbowanego Streszczenie Układem dynamicznym badanym w pracy jest podzespół zawierający podatną belkę poruszającą się ruchem obrotowym w płaszczyźnie poziomej. Zaprezentowano bezwymiarowe i perturbowane równania ruchu. Rozwiązania analityczne części zline- aryzowanej tych równań uzyskano dla przypadku idealnego i nieidealnego. Taka po- stać rozwiązania jest niezbędna do analizy zupełnego i słabo nieliniowego problemu, w którym nieliniowości stanowią niewielkie perturbacje wokół rozwiązania liniowego. Prezentowanewyniki badańmogą okazać się pomocą dla analityków zajmujących się modelowaniem układów dynamicznych uwzględniających interakcje zachodzące po- między daną konstrukcją bazową, a zamocowanymi na niej aktywnymi elementami wykonawczymi (aktuatorami). Manuscript received February 21, 2007; accepted for print July 16, 2007