Scribd
Upload
195 views

A Survey of The Dynamics Behaviour of A Simple Rotating Shaft With A Transverse Crack

Rotordynamics

Uploaded by

Joel Morales Perez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
195 views

A Survey of The Dynamics Behaviour of A Simple Rotating Shaft With A Transverse Crack

Rotordynamics

Uploaded by

Joel Morales Perez
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 20
Journal of Sound and Vibration (1993) 160(2), 313-332 A SURVEY OF THE DYNAMIC BEHAVIOUR OF A SIMPLE ROTATING SHAFT WITH A TRANSVERSE CRACK R. Gascu Technical University Berlin, Sekr. F5, Marchstrasse 12, D-1000 Berlin 10, Germany ‘(Received 2 January 1991, and in final form 29 August 1991) A simple but comprehensive survey is provided of the stability behaviour of a rotating shaft with a crack, and of the forced vibrations due to imbalance and to the crack. The analysis is restricted to that of the Laval rotor (one disk on an elastic shaft). Some possibili- ties for early crack detection are established. 1, INTRODUCTION The aim of this paper is to provide a simple, but comprehensive, survey of the stability behaviour of a rotating shaft with a crack, and of the force vibrations due to imbalance and to the crack, This is possible if vibrations remain small in comparison to the sag of the rotor under its own weight, Since the static sag of rotors threatening to crack is usually quite large, e.g., in turbo-generators often more than 1 mm, this approach is valid even when rotational speeds reach unstable ranges and the vibrations begin to grow. As long as the vibrations remain small in this sense, the non-linear equations of motion can be transformed to linear, periodically time-variant equations. ‘However, even given weight dominance in the displacement, a generally valid description in the sense of similarity theory is only possible in a few diagrams if the crack model is restricted to “strong” parameters. Only then are the results independent of the specific geometrical dimensions of the shaft. This is not the case in the analyses presented in references [1, 2]. The comparison of three different models of a cracked hollow shaft in reference [2] shows that, provided that the crack depth is not more than 50% of the shaft radius, the simple hinge model is very good for representation of the cyclic stiffness variables and for the stability limits. The same is true for a solid shaft. We therefore use a hinge model here, which only introduces one additional parameter, the additional (negative crack stiffness component ASzmax, which describes the main flex- ibility of the shaft when the crack is fully open. The cross-flexibility A5_,n0, Which becomes relevant only for very deep cracks, is a weak parameter and is omitted here (see Figure 3 of section 3). This paper is restricted to the analysis of the dynamics of the Laval rotor (one disk on an elastic shaft) with most simple crack model. A detailed discussion of nearly all crack models in use can be found in reference [3]. A brief general survey of the literature on cracked rotors was presented by Wauer in 1990 [4]. 313 (0022-460X /93/020313 +20 $08.00/0 © 1993 Academic Press Limited 314 R. GASCH 2. THE EQUATIONS OF MOTION IN AN INERTIAL FRAME AND THEIR LINEARIZATION TO PERIODICAL TIME-VARIANT CO-ORDINATES The equations of motion for a simple rotor with a cracked shaft have the following form in an internal frame [4-8]: m w\ [da fi) fon se (*:)-("") (or +20) [alll Mee IG) me cccan) © M #+ D &@+S@) u=p + BR Here w, and », are the displacements at the middle of the shaft in the plane of the disk of mass m (weight mg). Damping is d, and the eccentricity between the centre of gravity and the shaft centre is ¢; fis the angle between ¢ and the centre of the crack (see Figure 1). —e Figure 1. Simple rotor with oracked shaft. Inertial co-ordinates z-y and rotating co-ordinates {-n. Degrees of freedom wet, The stiffness matrix, which in the general case is non-linear and time-variant, can be split, S(u, )=S)+ AS(u, 1) Q) where Sp is the diagonal stiffness matrix of the uncracked shaft. AS is the additive (nega- tive) crack stiffness matrix, which is dependent both on the displacements u(¢) and on the rotation angle of the rotor y=.2t. However, a crack in a rotor usually causes a small change in its stiffness (i¢., AS is small compared to Sq). The vibrational response of the rotor can also be split, u(t) =u + Ants), @) where up, the static deflection of the uncracked shaft, is open ty) ‘ and Au(*) is the vector describing the vibrational behaviour. ‘CRACKED ROTATING SHAFT: A SURVEY 315 If weight dominance is assumed for the elastic deflection (Au(1)> Au(?), the weight deflection determines the opening and closing of the crack, and not the small vibrational movement Au(:): AS(u, 1) + AS(1). The crack stiffness AS(#), which is periodically variant because of y= :2t, governs the stability behaviour on the left side of equation (5) (homogeneous system). On the right side it governs the forced vibrations due to the crack through the expression —AS(1)0p. The stability of this system can be analyzed by using Floquet’s theory (see section 5), If stability is guaranteed, the forced vibrations can be determined from a time-invariant system MAu+ DAu+ SoAu=—AS(t)uy+p,. (6) because the term AS(z) Au is second order smallness. This is considered in section 6. Itis still assumed that weight effects dominate: i.e., that vibrational amplitudes are small ‘compared with the static weight deflection. This can easily be checked for a simple rotor. The weight deflection is Wesnas = M/Sos (7) where 59 is the stiffness of the uncracked shaft. However, for the eigenfrequency of the rotor, @3=59/m, so that one can write Westar = 8/0. (8) If the eigenfrequency of the uncracked rotor is known, one therefore knows the vibrational amplitudes for which the linearized theory is applicable (see Figure 2). In practice only narrow bands are excluded around the resonance peaks, 2/ao=1=4, } and 1. —r T 1158p = & fe 4 ¥ ost 4 o ° 100-200-300 wp(r0d/s) ° 1000 2000-3000 rpm) Figure 2. Static deflection Wea of the rotor as a function of the critical rotational speed (cigenfrequency).. 316 R. GASCH 3. THE CRACK MODEL 3.1. THE OPEN CRACK If a crack lies in the pressure zone of a shaft under load then it closes and the shaft has full flexural stiffness. However, if it lies in the zone subject to tensile forces the crack opens. The flexural stiffness of the shaft becomes less, as a result of the additional (local) flexibility (see Figure 3). One has we) _[ ot Al (é ("3 [ inn a ° For all but very deep cracks the cross-flexibility Af, is relatively unimportant compared to the main flexibility A’, as has been shown by calculations and experimental investiga tions. In Table 1 are shown values for the ratio Ak,/Altc, taken for a solid shaft from reference [9], and for a hollow shaft from reference [2]. Figure 3. Additional crack fe s, main Al and cross Alr,, for a fully open crack. TABLE 1 Ratio of cross to main crack flexibility versus crack depth ‘Ah,/Ahe 1/R=0 t/R=0-5 /R=1 3.2. THE BREATHING CRACK Ifa cracked shaft rotates slowly under the load of its own weight, then the crack will open and close once per revolution: it “breathes”. For a hollow shaft this process, which can be expressed in formulae developed from beam theory [2], is shown in Figure 4. ay is the (half) initial opening angle; @ is the current aperture angle as a function of the rotational angle y= 0 (a); static displacement W; n= in rotational co-ordinates ). From Figure 9 it is also possibie to read off directly the transformation matrix relating the rotational and the inertial co-ordinates: we\_[ eos 2¢ sin ar (") (") [s2, eal ») (a3) 4, = TO) a In order to determine the stiffness matrix in the inertial frame one first inverts equation (11) from flexibilities to stiffnesses in rotational co-ordinates, ()-E bol oD) fe = [So + AS(ro) Oe where then 1/(lig+ Ah) =59— Ass. (1s) For a real rotor, Asg and so can be determined experimentally, by determining the eigenfre- quency @» with a closed crack (crack uppermost) and then the eigenfrequency @, in the vertical direction for the fully open crack (crack down), whence (c¢/00)* As/s0- (16) CRACKED ROTATING SHAFT: A SURVEY 321 If the displacement u,,, in equation (14) is replaced by the transformation equation (13) and then multiplied from the left by T7 this gives the elastic forces in the inertial frame, f=T (1) fo (7) and thus the required stiffness matrix S(1) =Sp+ AS(1) =T"(So,o+ AS.) T. (18) Evaluated, this gives 2 F So _ cos’ Qt sin 21 cos Qt S+45() | al fi (05 ar sin? Q¢ | 9) or, slightly transformed, af Jia I+cos 201 sin 201 Sot 450) [ al af voas;|t 201 ~~ 1=cos20|" @) Tt can clearly be seen how the periodic time-variance of the steering function f(t) is modified by the transformation. 5. STABILITY ANALYSIS 5.1, FLOQUET’S METHOD The homogeneous, periodically time-variant system of equations of motion from equa- tion (5) with the matrices M and D as in equation (1) and the stiffness matrices as in equation (20) displays its stability properties during one period T:: that is, here, within one rotation of the rotor. For the stability investigation after Floquet we first have to calculate the transition matrix of the system (7), which tells us what happens to the initial condition x(¢=' Xp after one period: ‘Aw. ‘Aw,\ Av, dy, aif “POV ay en Aby, wT Ady, 10 x, =(T) %& This transition matrix can be obtained by first transforming the equations of motion to the state form and then integrating numerically over a period T. ‘When the transition matrix ®(7) has been obtained then after Floquet the proportional- ity between the variables of state initially (xo) and after the period 7(xr) is examined. Xr=UXo, (22) Substituting equation (22) in equation (21) then gives the Floquet eigenvalue problem which allows us to answer this question: [0(7)—p1}xo=0. (23) The vibrations die down when all the eigenvalues are less than 1; lel < 1. There is stability. If one or more eigenvalue is greater than | the system is unstable; |y_|> 1. The limiting stable case is then [yel= 1. 322 R. GASCH The factor |’ is easily interpreted: it shows the factor by which a given initial amplitude increases or decreases in the course of one revolution of the rotor. 5.2. RESULTS OF THE STABILITY ANALYSIS In Figure 10 is shown an overview of the stability behaviour of a cracked simple rotor obtained by using the hinge model, with no damping, D=0). The unstable ridges, |y1|>1 are clearly visible, beginning at the rotational speeds 2/o=2/l, 2/2, 2/3, 2/4,... Further narrow ridges at 2/5, etc., are suppressed by the graphics program. e Rotational speed, Q/y i ‘Figure 10. Stability of the cracked shaft against the “‘crack depth” and the dimensionless rotational speed, Damping D=0. Unstable zones |y1> 1. There is a broad unstable zone spreading out from 2/m)=2. The zone at 2/@o=1 corresponds to the one also known from the shaft with unequal moments of inertia (slots). A very unstable, but very narrow zone is clear at Q/w»=2/3, and an even narrower one at O/@=2/4. ‘The parameter used for the crack depth is the eigenfrequency a, of the “weakened” shaft over wo for the uncracked shaft (cf., equation (16). It should not be overlooked that the movements of the shaft in the unstable zone can increase extremely suddenly. As mentioned above, || is the factor by which the amplitudes increase during a single revolution of the shaft, Thus |4/=2 means that the amplitudes increase by a factor of 1024 in the course of 10 revolutions! It would also be an illusion to think that the lower values for || in the unstable zone at 2/e=2 are less threatening; this may be true for the period T (one revolution), but the shaft is turning twice as fast as at 2/@o=1, so that over time the increase is just as rapid. The stability limits for D= 1% and 5% were illustrated in Figure 8. Some damping eases the situation appreciably for shallow cracks, 6. FORCED VIBRATIONS If the stability of the system is ensured then instead of using equation (5) a good approximation of the forced vibrations can be achieved by using the time-invariant system {in equation (6). The expression AS(1) Au in equation (5) is then small and of the second CRACKED ROTATING SHAFT: A SURVEY 323 order. Introducing the stiffness matrix from equation (20) into equation (6) then gives m = Wui,\ [a Aw.) fo Yaw. [ olae}T Aas) alae) _ L+e0s201 sin201 fre cos (t+ B) =4s./00| sin 20¢ eel 0 )eemas( ee}, The system then behaves practically as a round shaft with a special excitation due to the crack AS(1)Up (the first term on the right) in addition to the excitation due to the imbalance (second term). ‘The following results are easily interpreted if the two deftection components Aw. and Av, are combined as a complex vector (see Figure 11) with an imaginary horizontal praxis: Ar=Aw,+jdv,, with j= JT. (24) Formally, the second of the system of equations of motion given above is multiplied by Rez Figure 11. Combination of the two displacement components Aw,, Av, into a complex vector. j= V7T and then both equations are added, with equation (24) taken into account. This gives an equation of motion of the form mAr + dAr + 507 = }Wes0dhse f(D +69] + ema? &O*P, (23) Once again the first term on the right side represents the periodic forced excitation due to the crack and the second the excitation due to the imbalance. 6.1. FORCED VIBRATIONS DUE TO IMBALANCE In order to demonstrate this complex notation, the vibrational response due to imbalance is examined, As;=0. With An = Ate (26) for equation (25) one immediately determines, from [-23mtjQd +t sJAF,= emQ?, (27) the imbalance response Are()= {17/1 1? +2)D,)} 1, (28) 324 R. GASCH where 17) =2/ag is the relative excitation frequency of the system and D=d/2map is the dimensionless damping of the system. At 1=1, when the rotational frequency @ equals the eigenfrequency of the shaft, the deflection Ar(1) is very large, tending to infinity as the damping tends to zero. In Figure 12(a) is shown the circular orbit with rotation in the same direction as the shaft (forward). Amplitude Of fe Rotational speed, = /arp oy (oy Figure 12. (a) Circular orbit of the imbalance response. (b) Imbatance amplitudes against dimensionless rotational speed. The amplitude of the imbalance response is shown in Figure 12(b). Details of this classic case can be found, for example, in reference {10} 6.2, FORCED VIBRATIONS DUE TO THE CRACK The periodic forced excitation due to the crack on the right side of equation (25) is proportional to the change in stiffiiess As; and the static deflection Wz: RHS.=LA5W sia SQ(1 +), (29) If the rectangular function in Figure 7 is written as a Fourier series one obtains S()=5+ (2/2) cos 1Qt— (2/32) cos 301+ (2/5) cos 5Qt (30) Then, by using the Euler formula cos nr=}(e""™ +e"), it is possible to express the crack-excitation force as kato ASH Tbe, Bl) koe R.HLS, where the participation factors 5; of the individual harmonics and their significance are shown in Table 2. The factors above the third order are either very small or exactly zero. ‘These participation factors are independent of the crack depth for the whole range 1/R<0-5 for which the hinge model is valid. ‘CRACKED ROTATING SHAFT: A SURVEY 325 TaBLe 2 Participation factors of the harmonics of the crack force excitation and their physical significance ‘Type of excitation by =4 Backward whirl Four times per revolution 0 -3 Three times per revolution 0-021 -2 Twice per revolution 0 -1 Once per revolution 0-106 0 Static, additional deflection 0-250 +1 Forward whirl Once per revolution 0318 42 Twice per revolution. 0-250 43 Three times per revolution 0-106 +4 Four times per revolution 0 The vibrational response of the simple shaft to the crack excitation is now easily deter- mined, From the differential equation with the crack excitation force component &, mAr +dAr + sehr = AseWepaibe O™, 32) with Ar = Ase one finds fg = AsWesrabe/ {EQ Jn + j(KD Jd +50} = Wesrud A S¢/50)bu/ {1 —K?2?/e03) + 25 DK(2/)}. (3) The overall response to the partial crack excitation by is = Ash beet Oza ey DE 4) where components of the fourth order and above can and have been neglected. The partial responses of the various harmonics will now be considered in more detail. For k™0, bo=0-25 one obtains the mean static additional deflection 0-25W-ssaAS¢/50, about which the oscillations of the shaft occur. ‘The partial components for K=~1, 6-1=0-106 and k= +1, b,,=0-318 are circular orbits, passed once per revolution of the shaft, with backward whirl (k=~1) and forward whirl (&= +1), respectively: As, 4. Fel & log’ FGjoyd° oO foo ie (on (35) Wepre oe ———__?*1__gt ian Sy 1 (2/0) —2j(Q/er0)D The superposing of the two orbits forms a narrow ellipse (see Figure 13) passed in the direction of rotation of the shaft (since |b4;|>|6-|). The resonance peak lies at Q/oo=1. 326 R. GASCH xa Byer . Ayettar Figure 13. The narrow ellipse passed once per revolution formed by the superposition of the forward and backward orbits (k= +1). The partial components for k= —2, b-2=O and k= +2, by.=0-25 deliver a simple circu- lar orbit, passed with double rotational frequency in the forward direction: — Ase baz on “(=04 Wy 8 —— ea 36 ArelO 04 Meat TG ]we?+ HQ/o—)D © & The resonance peak lies at 2/ay=1/2. The partial components for 3, b.5=—0-021 and k=+3, b4;=0:106 provide a slightly elliptical orbit, passed at triple rotational frequency, again in the forward direction since |b_3|<|b,3|. The resonance peak here lies at 2/a@o=1/3. These three circular or elliptical orbits are shown against the dimensionless rotational speed with damping, D=0-0S, in Figure 14. When these orbits are superimposed, taking into account the single, double or triple rotational frequency, the actual orbit is obtained. At the same time, however, the clarity of the situation is lost. This is demonstrated in Figure 15 at various rotational speeds 2/9. The point on the curve shows the sum of phases. In the experimental identification of cracks the Fourier spectrum will be recorded against the rotational speed. The spatial diagram obtained for the vertical amplitude of the rotating shaft is shown in Figure 16. In Figure 17 the (dimensionless) orbital radii are shown against the rotational speed (r.- to rs3). In particular, the magnification function for the backward whirl r, can provide important hints to the existence of a crack, as we shall see in the next section, 6,3, FORCED VIBRATIONS DUE TO THE CRACK AND THE IMBALANCE The overall response due to both excitations can be derived by combining the solutions from equations (34) and (28): Ar ()= Ar, + Atenee CRACKED ROTATING SHAFT: A SURVEY 327 Figure 14, Circular and elliptical orbits of single, double and triple rotational frequencies os. rotational speed. D=005. 0 o Oe 50 O28 on 040 oso O85 080 128 110 Figure 15. Orbits of the shaft centre as a resull of crack excitation. Rotational speed =0, 0-33, 0-40, etc. Damping D=0-05. 328 R. GASCH +o Rotational speed, 2/4 os 033) ol ° 05 1 15 Frequency, wo Figure 16, Fourier spectrum of the cracked shaft dependent on rotational speed; vertical amplitudes Aw, ; D=0-05. 03 05 1 oy ae i -1ar Os 1 Dhue Figure 17. The magnification function of the (a) forward and (b) backward circular orbits, which superimpose to give the orbit for the shat , D=0; —, D=0-05. The crack response components (single, double and triple rotational frequency) from Figure 15 are then superposed by the circular orbit of the imbalance response (see Figure 12(a)) which is also with forward whirl. Depending on the size em of the imbalance, the rotational speed @ and the angular position f, then the forward whirl component e™ of the crack response can be either ‘CRACKED ROTATING SHAFT: A SURVEY 329 reinforced or suppressed. This is immediately apparent if this component of the excitation force is written out from equation (27) and equation (31) or (25), Lem? e? + We sarhscbsi) ero, (38) or even better if this is transformed to a rotating co-ordinate system by multiplying by eo (see Figure 18). ematel® Reg Figure 18. Superposition of the excitation forces duc to crack and imbalance (both forward whirl), once per revolution. If one sets the balancing weight at B= 180° (e”=~1), 2m Q?=WeaaAscbs, (39) then the excitation due to the crack in the rotational direction at rotational speed can be fully compensated-—but only at the special speed (2. The resonance peak at =I is in general not eliminated, and may even be reinforced. If one wants to pass through this resonance of the cracked shaft without magnification then the treatment must be more sophisticated. The resonance has contributions from the imbalance and the crack: Ari(=ee® i (40) If, for simplicity’s sake, damping is set at D=0, then this can be rewritten as — 2 Ora to-[ee# v 1A btn HE eit I-¥ 5 2 -[(c 24D Wace 4s) ee | em, ap so l= 1 50 It is immediately clear that with B= 180° the resonance can be eliminated if em=mbWesi8e/505 (42) or, even easier to interpret, if Emad =b 4 Wer Sy (43) However, the rotor will still have a constant run-out once per revolution, bs Westar S¢/S0 as a result of the second term in equation (41). The backward whirl from k=—I cannot be compensated for. 330 R. GASCH However, such manipulations do not influence the stability, as long as Au0-5). We have already discussed the model developed by Mayes and Davies, on the basis of their experimental investigation of rotors with deep cracks. It includes the local cross- flexibility of the crack, which is not significant for 1/R<0-5 (see Table 1), and a modified steering function f(é) to reflect the fact that with deep cracks there is a smooth transition from opening and closing. This was already apparent for the theoretical analysis of the hollow shaft shown in Figure 4 for t/R=1 (a)=90°). For deep cracks the function (b) shown in Figure 7 is used: Alt)=(1 +008 20/2. (44) If this steering function is now used in equations (14) and (44) and also the additional cross-flexibility 4s,, then the coefficients b, of the forced vibration due to the crack are found in the form shown in Table 3. All higher terms not included in the table are zero. x is the cross-flexibility parameter (=As,/As-) which varies appreciably from zero only for the deeper cracks. Taste 3 Coefficients by of the forced vibraton due to a deep crack (t/ If the values of b, from Tables 2 and 3 are compared for x=0 then it is noticeable that the differences are very small except for the term 5,, with 0-318 against 0-25. It is also clearly shown in Table 3 that due to the influence of x as the crack depth increases the +2 and +3 terms become less significant. In general, the independence of the crack excitation term by from the crack depth, which exists for 1/R<0°5, is lost for deep cracks. 8, SUMMARY AND CONCLUSIONS The single parameter “hinge” crack model provides an overview of the dynamic behav- iour of a simple rotor with transverse crack in the range which is important for early detection, 1/R<0-5. CRACKED ROTATING SHAFT: A SURVEY 331 ‘An overview of the stabitity behaviour has been given in Figures 8 and 10. The orbits of forced vibrations due to the crack, the sizes of which are proportional to the static displacement of the rotor and the change in stiffness As; are quite complicated (see Figure 15). However, if these are split into components at single, double and triple rotational frequency, things become much clearer (see Figures 14, 16 and 17). Important indications of the existence of a crack can be obtained at rest from the measurement of the horizontal and vertical deflection of the rotor as a function of the angular position (see Figure 6). ‘When running the rotor up or down additional resonances at = 1/3 and 1/2 as well as an increase or decrease of the imbalance resonance at =I are signs of a crack, as are the marked phase migration which occur at these rotational speeds (cf., Figures 14 and 15). These have already been pointed out by Schmied [1]. It is most difficult to recognize while the rotor is rotating at constant speed. Here long- term observation of the mean static additional deflection can be useful. For rotors which are operated near the unstable zone at =2 a criterion has been developed in reference [12] of “increasing nervosity with increasing crack depth”. The recognition of cracks at rotational speeds above the critical range speed is so difficult because the significant double and triple frequency vibrations are then only very slightly involved in the crack response (cf., Figures 14-17). The +10 vibrations due to crack (41.9) and imbalance (+12) dominate. Here spectral analysis over long periods (trend analysis) can be helpful. If the crack grows (represented by As, in equation (34)) then all the 1.2, 2@ and 3@ amplitudes due to the crack grow to the same extent! If the imbalance remain virtually constant over this period then “today's spectrum minus yesterday's” can offer a clear indication of a crack— especially when taken together with the fact that (for isotropic bearings) the crack, in contrast to the imbalance, will excite a strong backward whirl component (—1{2 vibration) which cannot be balanced out. In addition to the references already cited, useful relevant material may also be found in references [13-15]. REFERENCES 1, J, Scuten 1986 VDI-Verlag, Reihe 11, Nr. 7, Diisseldorf. Schwingungsverhalten von Rotoren mit angerissenem Wellenquerschnitt, 2. R. Gascn, M. Person and B. Werrz 1988 Tec MECH Conference Vibrations in Rotating Machinery, September 1988, Edinburgh, UK, C314/88. Dynamic behaviour of the Laval rotor with a cracked hollow shaft—a comparison of crack models . W. Tuts 1990 VDI-Verlag, Reihe 11, Nr. 131 (139 pages) Disseldorf. Longitudinal and torsional vibrations in cracked rotors (in German). }. J. Wauer 1990 Applied Mechanical Reviews 43, 1. On the dynamics of cracked rotors: A Bterature survey. 9. R. Gascw 1975 IRL- Bericht 8, TU-Berlin (ISBN 3 7983 0551 X). Kleiner Beitrag zur Behandlung des dynamischen Verhaltens einer rotierenden Welle mit angerissenem Querschanitt. 6. 1. W. Mayes and W. G. R. Davies 1976 Vibrations in Rotating Machinery (Institution of Mechanical Engincers Conference Paper € 168/76). The vibrational behaviour of a rotating shaft system containing a transverse crack 7. R. A. Henry and B. E, OKAH-AVAE 1976 Vibrations in Rotating Machinery (Institute of Mechanical Engineers Conference Paper C 168/76). Vibrations in cracked shafts. 8, R. Gascu 1976 Vibrations in Rotating Machinery (Institute of Mechanical Engineers Conference Paper C 168/76). Dynamic behaviour of a simple rotor with a cross-sectional crack. 9. J. W. Mayes and W. G. R. Davies 1984 American Society of Mechanical Engineers, Journal of Vibration, Acoustics, Stress and Reliability in Design 106, 139-145. 10. R. Gascit and H. Prorrzver 1975 Rotordynamik——Eine Einfihrung. Berlin: Springer-Verlag. yee 332 R. GASCH 11. H-E. FREunp 1987 In Alianz-Berichte, Nr. 24, November 1987 “Schwingungstiberwachung von Turbosdtzen—ein Weg zur Erkennung von Wellenrissen”. Betriebserfahrungen bei einem Wel- lenriB einer 215 MW-Turbine, 12, E.E, Monte and H. Zrepanre 1987 In Allianz-Berichte, Nr. 24, November 1987 “Schwingungsii- berwachung von Turbosatzen—ein Weg zur Erkenrumg von Wellenrissen”. Fritherkennung von Querrissen an Turbinenwellen durch Schwingungsiiberwachung—Theorie und Anwendung, 13, T,INAGAKI, T. KANkrand K, SHIRAKI 1981 ASME Paper No. 81-Det-45, Transverse vibrations of a general cracked-rotor bearing system, 14, H. Zreparts, H. ScHwertrecer and E. E, MOnLe 1978 VDI-Berichte 320, 37-43. Auswir- kcungen von Querrissen auf das Schwingungsverhalten von Rotoren. 15, J, Mever 1979 Dissertation Hochschule der Bundeswehr Hamburg. Zum Schwingungsverhalten ciner Lavatwelle mit angerissenem Querschnitt.

You might also like