Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống
1
/ 40 trang
THÔNG TIN TÀI LIỆU
Thông tin cơ bản
Định dạng
Số trang
40
Dung lượng
628,72 KB
Nội dung
Chapter 1 Study on the Propagation of Inlet Flow Distortion in Axial Compressor Using an Integral Method An integral method is presented in this chapter. The integral method is proposed by Kim et al. at 1996 aimed to describe the problems of distorted inlet flow propagation in axial compressors. In this method, the detained blade rows are replaced by their equivalent force field, and the Navier-Stokes equation can be simplified by using in- tegral technique. Despite the great simplifications adopted in the integral method, the multistage axial compressor with large inlet distortion, including back flow can be analyzed and a qualitative trend of distortion propagation can be described successfully. The integral method provides the useful information about the per- formance of the axial compressor with inlet distortion, which is meaningful to engi- neering application in the design and analysis of turbomachinery. In this chapter, by applying the integral method, the effects of the parameters of inlet distortions on the trend of downstream flow feature in compressor calculated numerically is described. Other than the ratio of drag-to-lift coefficients of the blade and the angle of flow, the value of distorted inlet velocity is found to be another essential parameter to control the distortion propagation. Based on this calculation, a critical distortion line and corresponding distortion propagation fac- tor are proposed to express the effect of the two main inlet parameters: the angle of flow and the distorted inlet velocity, on the propagation of distortion. From the viewpoint of compressor efficiency, the distortion propagation is further de- scribed by a compressor critical performance. The results present a useful physical insight of compressor axial behavior and asymptotic behavior of the propagation of inlet distortion, and confirm the active role of compressor in determining the velocity distribution when compressor responds to an inlet flow distortion. 1.1 Introduction In the operation of aircraft jet-engine, it is important to understand the aerody- namic response resulting from an inlet flow distortion. An inlet flow distortion may cause the rotating stall or even surge, or a combination of them. If the blades fail to produce required loading, catastrophic damage to the complete engine would result. To avoid stall and surge of compressor due to flow distortion, and 2 understand further, the inlet distortion and its propagation effects have received much attention over the years. In 1955, Harry and Lubick [7] investigated the tur- bojet engine in an altitude chamber at the NACA Lewis laboratory to determine the effects of a wide range of uneven inlet-air pressure distributions on transient characteristics and stall phenomenon. The flow non-uniformities due to inlet distortion are commonly grouped into radially varying steady state, circumferentially varying steady state and unsteady distortions [5]. In many situations, the principal loss in stall margin can be regarded as due to one of the above three groups. Among them, the flow with circumferential distortion introduces new phenomenon into the fluid dynamic analysis of the com- pressor behavior and has gained much attention by many researchers. One of the methods used is a linearized approach, which provides quantitative information about the performance of the compressor in a circumferentially non-uniform flow. Several models, such as the parallel compressor model and its extensions, are used to assess the compressor stability with inlet distortion. The numerical solution of the time dependent nonlinear inviscid equations of motion is another method to study the problem of compressor stability in a distorted flow. In 1980, Stenning [11] presented some simpler techniques for analyzing the ef- fects of circumferential inlet distortion. Stenning concluded that it is impossible to achieve a complete success in calculating the distorted performance and distortion attenuation of an axial compressor even though the use of high speed computers has greatly improved the accuracy of performance analysis somewhat. However, as an attempt, it is necessary to study the distorted performance and distortion at- tenuation of an axial compressor even though a complete success may not be achievable. As a first approach, the methods that use the known undistorted per- formance characteristic as a starting point to predict the behavior of the compres- sor with distortion are developed with a much prospect for success. A number of investigators ([2], [3], [5], [6], [10]) have developed models for response to circum- ferential distortion which use the undistorted compressor performance or stage characteristics to predict the behavior with distortion. One novel real time correla- tion scheme for the detection of operations (both steady state and transient) near to the stability boundary of a compressor has recently been studied at the Georgia In- stitute of Technology [9]. The information provided by the correlation scheme is stochastic in nature to facilitate controller design [4]. The scheme has been linked to the engine fuel controller and was used detect and avoid impending stall by modulating the transient fuel schedule during operability tests [1]. Kim et al. [8] also successfully calculated the qualitative trend of distorted per- formance and distortion attenuation of an axial compressor by using a simple inte- gral method. This simple integral method was applied to describe the qualitative trend of distortion propagation in axial compressors. Kim et al. concluded that the two key parameters to control the distortion propagation are the ratio of drag-to- lift coefficients of the blade and the angle of flow. The integral method pro- vides the useful physical insight and qualitative information about the perform- ance of the axial compressor with inlet distortion. It is meaningful to make use of and develop this method further. Chapter 1 Study on the Propagation of Inlet Flow Distortion 3 In this chapter, the integral method is confirmed and further improved. An in- vestigation is progressed to understand the downstream flow feature with the inlet distortion, including the velocity and mass flow rate. A critical distortion line is presented to express the effect of two essential inlet parameters on the propagation of distortion. The critical distortion line is simple, efficient and complete expres- sion to analyze the propagation of inlet distortion in axial compressor. Finally, the compressor critical performance and critical characteristic are also discussed. 1.2 Theoretical Formulation Consider a two-dimensional inviscid flow through an axial compressor as shown in [8]. The computational domain is shown in Fig. 1.1. The flow is described by equations of continuity and motion: 0= ∂ ∂ + ∂ ∂ y v x u (1.1) x F p x uv y u x = ∂ ∂ + ∂ ∂ + ∂ ∂ )()()( 2 ρ (1.2) y F p y v y uv x = ∂ ∂ + ∂ ∂ + ∂ ∂ )()()( 2 ρ (1.3) distorted inlet undistorted inlet v, y u, x δ δ (Y (x) y=Y (x)+2 R 2 1 +Y (x)) 1 π y=Y(x) =1/2 y=Y (x) 2 y=Y (x) 1 2R π distorted inlet undistorted inlet v, u, x 1 1 2R π η=1 2R π η=−1 η=0 η= /δ−1 /δ η coordinates transfer from (x, y) to (x,η) : Fig. 1.1. A schematic of coordinates transfer of calculation domain 1.2 Theoretical Formulation The coordinate system is transferred from (x, y) to (x, η ) by: δ η )x(Yy − = (1.4) here )x(Yy = is a symmetric line of distorted region, and δ is a half height of distorted region. If the static pressure takes circumferentially uniform, then: )x( pp ρρ ≡ (1.5) Before integrating the x- and y- momentum equations, and mass conservation equation in distorted region, undistorted region and overall region, respectively, some assumptions and parameters definitions are given for velocity, pressure and forces. 1.2.1 Velocity and Pressure Assume the inlet velocity has a windward angle of 0 θ , then 000 tan V U γ θ == (1.6) where 0 U and 0 V are the x- and y- components of inlet velocity, respectively. α and β are the x- and y-velocity increments in the distorted inlet region, and 0 α and 0 β are the x- and y-velocity increments in the undistorted inlet region, respectively. In distorted region: 0 Uu α = (1.7) 0 Vv β = (1.8) In undistorted region: 00 Uu α = (1.9) 00 Vv β = (1.10) Chapter 1 Study on the Propagation of Inlet Flow Distortion 4 2 0 )V2(21 p P ρ = (1.11) The non-dimensionalized velocity parameter (rotor speed) is 0 V2R = ω . The pressure is non-dimensionalized by 2 0 )V2( 2 1 ρ : 1.2.2 Forces ( )( ) *22 L S tantanvu 2 1 kF θθ −+= ⊥ (1.12) ( )( ) 2 *22 D S II tantanvu 2 1 kF θθ −+= (1.13) () [ ] () * 2 2 L R tantanvRu 2 1 kF ψψω −−+= ⊥ (1.14) () [ ] () 2 * 2 2 D R II tantanvRu 2 1 kF ψψω −−+= (1.15) The force components for a unit circumferential distance of the entire blade row are: () ( ) θθ λ λ ψψ λ λ cosFsinFcosFsinFF S II S S R II R R x −+−= ⊥⊥ (1.16) () ( ) θθ λ λ ψψ λ λ sinFcosFsinFcosFF S II S S R II R R y −−++= ⊥⊥ (1.17) 1.2 Theoretical Formulation 5 As shown in Fig. 1.2, in integral equations, x F and y F denote the forces in the dis- torted inlet region, and 0,x F and 0,y F denote the forces in the undistorted inlet region, respectively. F θ v u F ⊥ F // F ⊥ F ⊥ ⊥ F // F // F // λ s λ r R OT O R ST A TO R θ r s u ω R -v Fig. 1.2. The force diagram and velocities in the compressor stage { () [ ] () () [ ] ( ) } { ()( ) ()( ) } θθθθθθ λ λ ψψψωψψψω λ λ costantanvu 2 k sintantanvu 2 k costantanvRu 2 k sintantanvRu 2 k F 2 *22 D *22 L S 2 * 2 2 D * 2 2 LR x −+−−++ −−+−−−+= () [ ] [ ] 2 22 2 * RL 00 *2 D 222 L 22 22 * *2 S LD 00 222 L k22 U2 V( tan) 2 k2 ( tan ) k (2 ) kk (U V)( tan ) ( tan ) 2k λββ αβ γψγ λαα βα γψ α αβγ λ βββ α αβ γθγ γθ λααα α βγ −− ⎡⎤ =+− − ⎣⎦ − −− +− ++−−− + (1.20) { () [ ] () () [ ] ( ) } { ()( ) ()( ) } θθθθθθ λ λ ψψψωψψψω λ λ sintantanvu 2 k costantanvu 2 k sintantanvRu 2 k costantanvRu 2 k F 2 *22 D *22 L S 2 * 2 2 D * 2 2 LR y −+−−+−+ −−++−−+= () [ ] [ ] 2 22 2 * RL 00 *2 D 222 L 22 22 * *2 S LD 00 222 L k2 U2 V( tan) 2 k2 2 ( tan ) k (2 ) kk (U V)( tan ) ( tan ) 2k λβ αβ γψ λα ββα γψ γ αα αβγ λ βββα αβ γθ γθγ λααα α βγ − ⎡⎤ =+− − ⎣⎦ −− +− +− −+−+− + (1.21) In undistorted region, by substituting (1.6), (1.9), (1.10), (1.20), (1.21) and (1.12), (1.13), (1.14), (1.15) into (1.16) and (1.17), using similar procedure as in the dis- torted region, we obtain: Chapter 1 Study on the Propagation of Inlet Flow Distortion 6 Assuming 00 VVR =− ω , in distorted region, by substituting (1.6), (1.7), (1.8), (1.18), (1.19) and (1.12), (1.13), (1.14), (1.15) into (1.16) and (1.17), we obtain: where the superscript S and R denote the stator and rotor, respectively; the λλ R and λλ S are generally functions of x ; L k and D k are lift and drag coefficients, respectively. The angles θ and ψ are the functions of the local velocity compo- nents and the parameter ω : uvtan = θ (1.18) () uvRtan −= ωψ (1.19) 7 () [ ] [ ] 2 22 2 * 0 RL y,0 0 0 0 0 0 *2 000 D 222 L0 0 00 22 22 * *2 S0000 LD 00 00 222 0L00 00 2 k FU2V(tan) 2 22 k (tan) k (2 ) kk (U V)( tan) ( tan) 2k β λ αβ γψ λα ββα γψ γ αα αβγ λβββα α β γθ γθ γ λααα α βγ − ⎡⎤ =+− −+ ⎣⎦ −− −− +− +−+− + (1.23) Now, the (1.1), (1.2), (1.3) can be integrated in both distorted region and undis- torted region. 1.2.3 Distorted Region . 1 1 constud = ∫ − ηδ (1.24) [][ ] ∫∫∫ −−− = ∂ ∂ ′ + ′ −+ 1 1 1 1 1 1 2 )( ηδη ρη ηδδηδ dFd p Y dx d du dx d x (1.25) [] ∫∫ −− =−−+ 1 1 1 1 )1,()1,( ηδ ρρ ηδ dFx p x p uvd dx d y (1.26) 1 K R ≡ π δα (1.27) Then from (1.25): [] ] ∫∫∫ −−− =+ 1 1 1 1 1 1 2 0 2 )( ηδη ρ δηαδ dFd p dx d dU dx d x (1.28) 1.2 Theoretical Formulation Accounting for the fact that the boundaries are streamlines, [8] developed the integral expressions as: From (1.24) and (1.7), we can obtain a constant δα . We introduce a constant K 1 and set: () [ ] [ ] 2 22 2 * 00 RL x,0 0 0 0 0 00 *2 00 D 222 L0 00 22 22 * *2 S0000 LD 00 00 222 00L0 00 22 k FU2V(tan) 2 2 k (tan) k (2 ) kk (U V)( tan) ( tan) 2k β β λ αβ γψγ λαα βα γψ α αβγ λβββα αβ γθγ γθ λααα α βγ − − ⎡⎤ =+− − − ⎣⎦ − −+ +− +−−− + (1.22) , and then: y 2 0 F d1 () dx U β α γ = (1.31) 1.2.4 Undistorted Region Similarly, in undistorted region, integrating the (1.1), (1.2), (1.3) in [1, 1 2 − δ π R ], we obtain: . 1 2 1 constud R = ∫ − δ π ηδ (1.32) [ ][ ] ∫∫∫ −−− = ∂ ∂ ′ + ′ −+ 1 2 1 0, 1 2 1 1 2 1 2 )( δ π δ π δ π ηδη ρη ηδδηδ R x RR dFd p Y dx d du dx d (1.33) [ ] ∫∫ −− =−−+ 1 2 1 0, 1 2 1 )1,()1 2 ,( δ π δ π ηδ ρδ π ρ ηδ R y R dFx pR x p uvd dx d (1.34) From (1.33): [ ] ] ∫∫∫ −−− =+ 1 2 1 0, 1 2 1 1 2 1 2 0 2 0 )( δ π δ π δ π ηδη ρ δηαδ R x RR dFd p dx d dU dx d (1.35) or: 22 0x,0 2 0 ddP1 ()2 F dx dx U αγ += (1.36) Similarly, for the y –momentum equation: [ ] ∫∫ −− = 1 2 1 0, 1 2 1 0000 δ π δ π ηδηβαδ R y R dFdVU d x d (1.37) Chapter 1 Study on the Propagation of Inlet Flow Distortion 8 2 x 2 0 ddP1 2F dx dx U α αγ += (1.29) [] ∫∫ −− = 1 1 1 1 00 ηδηαβδ dFdVU dx d y (1.30) Using (1.6), (1.11) and (1.27) and from (1.25) and (1.5), we can obtain: Similarly, the y –momentum equation can be deducted from (1.26) plus the as- sumption of constant circumferential static pressure: 9 1.2.5 Entire Region Integrating the overall conservation: . 1 2 1 00 1 1 0 constdUdU R ≡+ ∫∫ − − δ π ηαδηαδ (1.39) yield: [] 00000 )2())(22(2 KRUURU ≡ − + π α δ π α δ (1.40) Using (1.27): 0011 )1( KKK =++ α α (1.41) Rearranged as: 1 10 0 )( K KK − − = α α α (1.42) The overall x- momentum equation is: [ ] ][ ] ∫∫∫∫∫ − − − − − − +=++ 1 2 1 0, 1 1 1 2 1 1 2 1 2 0 2 0 1 1 2 0 2 )( δ π δ π δ π ηηδη ρ δηαηαδ R xx RR dFdFd p dx d dUdU dx d (1.43) Integrated as: )( 1 0, 2 0 2 0 xx FF Udx d dx d −=− α α α (1.44) Differentiating (1.42) results in: dx d K dx d α α 2 0 = (1.45) Where: 2 1 011 2 )( )( K KKK K − − = α (1.46) 1.2 Theoretical Formulation or: y ,0 00 2 0 F d1 ()() dx U αβ γ = (1.38) dx d K KKK dx d dx d α α α α α α 3 1 2 0110 0 2 0 )( )(2 2 − − −== (1.47) The left-hand-side of (1.44) becomes: dx d K dx d dx d α α α α α 3 2 0 =− (1.48) where: 3 1 2 011 3 )( )(2 1 K KKK K − − += α (1.49) Substituting above equation into (1.44), yield: )( 1 0, 2 03 xx FF UKdx d −= α α (1.50) 1.2.6 Integral Equations From above integrated results, we can find out five ordinary differential equations which describe the progress of the distorted and undistorted regions, as well as the progress of pressure as flow moves downstream in the compressor. The first one is (1.50), the second one is (1.31), and the third one is (1.45). From the (1.45), we obtain: y,0 00 00 2 0 F dd 1 () dx U dx β α αβ γ =− (1.51) Then, using (1.45), we obtain the fourth integral equation. From (1.29), the fifth integral equation can be derived. Putting all of the five integral equations together, there are: x x,0 2 30 FF d1 () dx K U α α − = (1.52) y 2 0 F d1 () dx U β α γ = (1.53) 0 2 d d K dx dx α α = (1.54) Chapter 1 Study on the Propagation of Inlet Flow Distortion 10 Using (1.42) and (1.45): [...]... Pi ) (1. 71) K 21 = Δx f1 ( xi + K K K K K Δx ,α i + 11 , β i + 12 ,α 0 i + 13 , β 0 i + 14 , Pi + 15 ) 2 2 2 2 2 2 (1. 72) K 22 = Δx f 2 ( xi + K K K K K Δx ,α i + 11 , β i + 12 ,α 0 i + 13 , β 0 i + 14 , Pi + 15 ) 2 2 2 2 2 2 (1. 73) K 23 = Δx f 3 ( xi + K K K K K Δx ,α i + 11 , β i + 12 ,α 0 i + 13 , β 0 i + 14 , Pi + 15 ) 2 2 2 2 2 2 (1. 74) K 24 = Δx f 4 ( xi + K K K K K Δx ,α i + 11 , β i + 12 ,α... + 12 ,α 0 i + 13 , β 0 i + 14 , Pi + 15 ) 2 2 2 2 2 2 (1. 75) K 25 = Δx f 5 ( xi + K K K K K Δx ,α i + 11 , β i + 12 ,α 0 i + 13 , β 0 i + 14 , Pi + 15 ) 2 2 2 2 2 2 (1. 76) K 31 = Δx f1 ( xi + K K K K K Δx ,α i + 21 , β i + 22 ,α 0 i + 23 , β 0 i + 24 , Pi + 25 ) 2 2 2 2 2 2 (1. 77) K 32 = Δx f 2 ( xi + K K K K K Δx ,α i + 21 , β i + 22 ,α 0 i + 23 , β 0 i + 24 , Pi + 25 ) 2 2 2 2 2 2 (1. 78) K 33 = Δx... (1. 60) F dP dα 1 = ( x2 − α = f 5 ( x,α , β ,α 0 , β0 ,P ) ) dx U0 dx 2 2 (1. 61) 1. 3 .2 4th-order Runge-Kutta Equations 1 ( K 11 + 2 K 21 + 2 K 31 + K 41 ) 6 (1. 62) β i +1 = β i + 1 ( K 12 + 2 K 22 + 2 K 32 + K 42 ) (1. 63) α 0 i +1 = α 0 i + 1 ( K 13 + 2 K 23 + 2 K 33 + K 43 ) (1. 64) β 0 i +1 = β 0 i + 1 ( K 14 + 2 K 24 + 2 K 34 + K 44 ) (1. 65) α i +1 = α i + 6 6 6 Pi +1 = Pi + 1 ( K15 + 2 K 25 + 2. .. i + 21 , β i + 22 ,α 0 i + 23 , β 0 i + 24 , Pi + 25 ) 2 2 2 2 2 2 (1. 79) K 34 = Δx f 4 ( xi + K K K K K Δx ,α i + 21 , β i + 22 ,α 0 i + 23 , β 0 i + 24 , Pi + 25 ) 2 2 2 2 2 2 (1. 80) K 35 = Δx f 5 ( xi + K K K K K Δx ,α i + 21 , β i + 22 ,α 0 i + 23 , β 0 i + 24 , Pi + 25 ) 2 2 2 2 2 2 (1. 81) K 41 = Δx f1 ( xi + Δx,α i + K 31 , β i + K 32 ,α 0 i + K 33 , β 0 i + K 34 , Pi + K 35 ) (1. 82) K 42 = Δx... DU2=SQRT(ALFA (1, I)* *2+ (ALFA (2, I)*GAMA)* *2) *Y(I) DU=(DU1+DU2)/(GAMA *2) DU0=SQRT(ALFA(3,I)* *2+ (ALFA(4,I)*GAMA)* *2) /(GAMA *2) DP 12= ALFA(5,I) A1=SQRT(ALFA (1, I)* *2+ (ALFA (2, I)*GAMA)* *2) /(GAMA *2) WRITE(*,*)ALFA (1, 1),SITA1,DY1 WRITE (1, 21 0 )DU0,DP 12, DU,DM,DP0,A1,ALFA (1, 1),SITA1,DY1 SI0=SI0+DSI0 SITA1=SI0 IF (SITA1.GT .25 .) GOTO 22 0 DY=0 .1 GOTO 1 END IF AL1=ALFA (1, 1) DY=Y(I)-Y (1) 10 9 IF (ALFA (1, 1).LT .1. 01) THEN AL0=AL0+DAL0 ALFA (1, 1)=AL0... (1) DISTORTED REGION: FX1=ALFA (1, I -1) * *2+ ( (2. -ALFA (2, I -1) )*GAMA)* *2 FX2= (2. -ALFA (2, I -1) )*GAMA/ALFA (1, I -1) -TAN(PSAIS) FX3= (2. -ALFA (2, I -1) )*GAMA/ALFA (1, I -1) * (1. -DL)+ & DL*TAN(PSAIS) FX4=ALFA (1, I -1) /(SQRT(ALFA (1, I -1) * *2+ & ( (2. -ALFA (2, I -1) )*GAMA)* *2) ) FX5=ALFA (1, I -1) * *2+ (ALFA (2, I -1) *GAMA)* *2 FX6=ALFA (2, I -1) *GAMA/ALFA (1, I -1) -TAN(SITAS) FX7=ALFA (2, I -1) *GAMA/ALFA (1, I -1) * (1. -DL)+ & DL*TAN(PSAIS) FX8=ALFA (1, I -1) /(SQRT(ALFA (1, I -1) * *2+ ... FX8=ALFA (1, I -1) /(SQRT(ALFA (1, I -1) * *2+ & (ALFA (2, I -1) *GAMA)* *2) ) Appendix 1. A Fortran Program: Integral Method 31 FXU(I)=SKL/4.*(FX1*FX2*FX3*FX4+FX5*FX6*FX7*FX8) ! FX/U0* *2 FY1 =1. +FX2*DL* (2. -ALFA (2, I -1) )*GAMA/ALFA (1, I -1) FY2 =1. +FX6*DL*ALFA (2, I -1) *GAMA/ALFA (1, I -1) FYU(I)=SKL/4.*(FX1*FX2*FY1*FX4-FX5*FX6*FY2*FX8) ! FY/U0* *2 ****** (2) UNDISTORTED REGION: FX 01= ALFA(3,I -1) * *2+ ( (2. -ALFA(4,I -1) )*GAMA)* *2 FX 02= (2. -ALFA(4,I -1) )*GAMA/ALFA(3,I -1) -TAN(PSAIS)... γ − ζ )[ γ ( 1 − D ) + D ζ α α 4 kL kL + ] α α + ( 2 − β )2 γ 2 2 α ] α 2 + β 2 2 (1. 87) Fy = kL 2 ⎡ 2 2−β k 2 β 2 β 2 U 0 α + ( 2 − β ) γ 2 ]( γ − ζ )[ 1 + D ( γ −ζ ) γ ⎣ 4 α kL α α − kL 2 2 k β β β U 0 ( α + β 2 2 )( γ − ζ )[ 1 + D ( γ − ζ ) γ 4 kL α α α ] α α 2 + ( 2 − β )2 γ 2 α ] α + β 2 2 2 (1. 88) Fx ,0 = + 2 − β0 2 − β0 kL 2 k k 2 U 0 [ α 0 2 + ( 2 − β0 ) γ 2 ]( γ − ζ )[ γ (1 D )+ D ζ 4... F (1, I)=(FXU(I)-FX0U(I))/(CK3*A (1) ) F (2, I)=FYU(I)/(GAMA*A (1) ) F(3,I)=CK2*F (1, I) F(4,I)=FY0U(I)/(GAMA*A(3))-A(4)/A(3)*CK2*F (1, I) F(5,I)=(FXU(I)-F (1, I))/(GAMA* *2* 2) DO 2 01 K =1, 5 2 01 RK (2, K,I)=DX*F(K,I) ! K 21 , K 22, K23,K24,K25 DO 300 K =1, 5 300 A(K)=ALFA(K,I -1) +RK (2, K,I) /2 F (1, I)=(FXU(I)-FX0U(I))/(CK3*A (1) ) F (2, I)=FYU(I)/(GAMA*A (1) ) F(3,I)=CK2*F (1, I) F(4,I)=FY0U(I)/(GAMA*A(3))-A(4)/A(3)*CK2*F (1, I) F(5,I)=(FXU(I)-F (1, I))/(GAMA* *2* 2)... growing 1. 4 Results and Discussion 0.0 0 .2 0.4 0.6 0.8 5 23 1. 0 6 24 24 22 22 20 distortion growing Δξ>0 18 θ ° 18 20 4 16 16 1 14 2 3 14 0 Δξ= 12 10 0.0 12 distortion lessening Δξ . () 413 1 21 1 11 22 6 1 KKKK ii ++++= + αα (1. 62) () 423 22 2 12 i1i KK2K2K 6 1 ++++= + ββ (1. 63) () 433 32 313 i 0 1i 0 KK2K2K 6 1 ++++= + αα (1. 64) () 443 42 414 i 0 1i 0 KK2K2K 6 1 ++++= + ββ (1. 65). ) } θθθθθθ λ λ ψψψωψψψω λ λ costantanvu 2 k sintantanvu 2 k costantanvRu 2 k sintantanvRu 2 k F 2 *22 D *22 L S 2 * 2 2 D * 2 2 LR x −+−−++ −−+−−−+= () [ ] [ ] 2 22 2 * RL 00 *2 D 22 2 L 22 22 * *2 S LD 00 22 2 L k 22 U2 V( tan) 2 k2 ( tan ) k (2 ) kk (U V)( tan ) ( tan ) 2k λββ αβ. ) 2 , 2 , 2 , 2 , 2 , 2 ( 15 14 0 13 0 12 11 1 21 K P K K KK x xfxK i ii iii +++++ Δ +Δ= βαβα (1. 72) ) 2 , 2 , 2 , 2 , 2 , 2 ( 15 14 0 13 0 12 11 22 2 K P K K KK x xfxK i ii iii +++++ Δ +Δ= βαβα (1. 73) ) 2 , 2 , 2 , 2 , 2 , 2 ( 15 14 0 13 0 12 11 323 K P K K KK x xfxK i ii iii +++++ Δ +Δ= βαβα