Response of a Squeeze Film Damper-Elastic Structure System to Multiple and Consecutive Impact Loads

Transcription

1 Proceedings of ASME Turbo Expo 2016: Turbine Technical Conference and Exposition, June 13-17, 2016, Seoul, South Korea Paper GT Response of a Squeeze Film Damper-Elastic Structure System to Multiple and Consecutive Impact Loads Luis San Andrés Mast-Childs Chair Professor, Associate Director of Turbomachinery Laboratory, ASME Fellow Sung-Hwa Jeung Graduate Research Assistant, ASME Member Supported by Pratt & Whitney (UTC) 1

2 Objective Impact Aircraft engines endure sudden maneuver loads (hard landing, etc.) t i Time Conduct impact load tests for motions starting from centered and off-centered positions evaluate SFD dynamic forced performance and identify system damping ratio. 2

3 Brief literature review Fluid inertia effect under an impulsive load Tichy and Bou-Said (1991) A numerical case study: fluid inertia force leads to reduction in journal peak amplitude of motion. Lee et al. (2006) Transient response of a rotor-bearing system is sensitive to the time duration of the external shock. San Andrés and Jeung (2015) SFD impact load tests: (amplitude motion/load amplitude)~ constant system damping ratio increases with amplitude. 3

4 SFD Test Rig cross sections Test Journal Bearing Cartridge Piston ring seal (location) Supply orifices (3) Flexural Rod (4, 8, 12) Main support rod (4) Journal Base Pedestal in

5 SFD Test Rig cut section L/D=0.2 Short length SFD Geometry (three feed holes 120 o apart) Journal Diameter, D 12.7 cm (5.0 in) Land Length, L 2.54 cm (1.0 in) Radial Land Clearance, c Feed orifice Diameter, ϕ 2.54 mm (0.1 inch) 267 μm (10.5 mil) Re s =(r/m)w c 2 = 27 at w =200 Hz 5

6 Impact Transient response due to a single impact load Time 6

7 Tests conducted Impact load tests for motions from centered and off-centered static positions. (e s /c=0 to 0.6) Y ϴ=90 Z STAT =(e S )cos(45º) ϴ=45 F S Z MAX = Z STAT + Z DYN Z dyn MAX 11 n dyn Z 1 MAX c c n i i Journal e S F X ϴ=0 X FMAX 1 1 L D L D n n i1 F MAX Number of impact i n 20 ϴ=225 BC Max. load amplitude, Peak amplitude, Static eccentricity, Motion Type F MAX /(LD) (bar) Z DYN /c (-) e s /c (-) Unidirectional

8 Impact load and bearing displacement Motion is transient and decays fast. 8

9 Impact load and bearing displacement F Z dyn MAX MAX c ( L D) Static eccentricity increases Peak BC displacement is proportional (linear) to magnitude of impact load. 9

10 Damping ratio and log dec Response of a viscous underdamped system wn t d d Z( t) e ( Acosw t Bsin w t) w n K M S 2 C KM S w w 1 d n 2 Logarithmic decrement M M M BC SFD

11 (a) Z X (t) (b) Z X (ω) Transient response damping ratio The higher the initial static eccentricity and peak load, the fastest the decay. system damping ratio increases. 11

12 Damping ratio & log dec vs peak motion Damping ratio ξ [-] (0.2) cos(45º)=0.14 (0.4) cos(45º)=0.28 (0.6) cos(45º)=0.42 Log dec δ [-] Z MAX = Z STAT + Z DYN e s /c=0.6, Y e s /c=0.4, Y e s /c=0.2, Y e s /c=0, Y e s /c=0.6, X e s /c=0.4, X e s /c=0.2, X e s /c=0, X Maximum BC displacement, [Z MAX /c] Damping ratio increases both with the BC amplitude (Z MAX ) and the initial static eccentricity e s Y F s F X X F Y 12

13 Compare damping from two configurations STLE (2015) Current Clearance: c s =213 μm (8.4 mil) c=267 μm (10.5 mil) 25% larger film clearance 13

14 System damping ratio motions around center 2 2 s c 267 ~ 1.57 cs 213 Damping ratio () for the small film clearance (c s ) is ~1.3 to 1.6 more than for larger clearance (c) SFD. 14

15 Conclusions(1) GT Transient response due to a single impact load (a)the SFD transient response decays faster when applied load (F) is large; the more off-centered (e s ), the larger the decay. (b) System damping ratio (ξ) increases linearly with peak displacement Z MAX. (b)an identical damper with smaller clearance shows ξ = ξ of current damper (c=1.25c s ) 15

16 Impact Transient response due to multiple impact loads t i Time 16

17 Sequences of impacts with equal load amplitude Impact Increase number of impacts. Motions from centered (e s =0). t i Number of impacts Duration of impact (ms) Time Time between impacts, t i (ms) Max. load amplitude, F MAX /(LD) (bar) 2, 3, 4,

18 Case 6: Six impacts t i = 30ms e S /c=0.0 Peak displacement vs Impact load Motions about e s =0 (Insets show time traces of impact load and ensuing BC displacement). Six consecutive impacts with elapsed time between impacts t i ~30ms. F Z dyn MAX MAX c ( L D) [1/bar] = slope of line fit to data 18

19 Peak motion amplitude vs # of impacts Impact Sequences of impacts (t i =30 ms) with same load magnitude ~0.07 1/bar t i Time e S /c=0.0 t i = 30ms F Z dyn MAX ( L D) MAXFindings: β increases slightly as number of impacts increases. c 19

20 Consecutive Impacts with decaying load Impact F MAX Consecutive impacts with decreasing load for motions from static eccentricity e s /c=0.0, 0.2, 0.4, and 0.6. Load halves on next impact. Time Number of consecutive impacts Duration of impact (ms) Time between impacts, t i (ms) Max. load amplitude, F MAX /(LD) (bar) Static eccentricity e s /c 1, 3, , 0.2, 0.4,

21 Impact load bearing displacement Consecutive impacts with load halving. 1 Imp Motions departing from e s =0 and F MAX /(LD)=7.9 bar 3 Imps 4 Imps 21

22 BC velocity and displacement Findings: At the incidence of an impact (circled in black), the impulse load adds velocity. Z t Z t F dt 0 M BC 22

23 Peak displacement vs Impact load Motions about e s = 0 & 0.4 Tests with 3 and 4 consecutive impacts and t i ~0ms. F Z dyn MAX MAX c ( L D) [1/bar] = slope of line fit to data 23

24 Peak motion amplitude vs # of impacts Consecutive impacts with load halving t i = 0ms X direction ~0.07 1/bar F Z dyn MAX MAX c ( L D) β rises as number of applied consecutive impacts increases. Little influence of static eccentricity. 24

25 Comparisons of experimental force coefficients to predictions from a physical model 25

26 Model SFD with fluid inertia Equation of motion for test system as a point mass Y M a C Z K Z F F F X SX X SX X X() t SX SFD X BC a F Y CSY ZY KSY ZY Y() t F SY F SFDY housing Support structure Applied force Axially averaged pressure field (P) (with fluid inertia) SFD F X X 2 2 rh ( cos sin ) ( cos sin ) 3 X Y X Y X cos Y sin ml P Z Z Z Z Z Z h 12m h SFD force F SFD f ( Z, Z, Z, Z, Z a, Z a ) X Y X Y X X Y Y F cos RL P d X 2 0 ( ) F Y sin SFD Integrate numerically to find response Z(t) 26

27 Predictions: time response Re s =(r/m)w c 2 = 27 at w =200 Hz Findings: Fluid inertia in the squeeze film greatly attenuates the increase of peak displacement well known from Tichy and Bou-Said (1991) 27

28 β: Predictions & experimental results F Z dyn MAX MAX c ( L D) Predictions including fluid inertia agree with experimental results. t i = 30ms e S /c=0.0 During a transient event, fluid inertia in the squeeze film reduces the peak response (lowers β). 28

29 Conclusion (2) GT Transient response due to multiple impact loads (a) For a sequence of impacts (t i =30ms), peak Z MAX is proportional to F MAX (b) For a series of consecutive impacts (t i =0ms), the peak amplitude increases with an increase in the number of impacts. (c) Fluid inertia reduces the amplification of peak displacement. (d) Neglecting fluid inertia in the prediction of SFD forces leads to motions ~50% larger in amplitude than those measured. 29

30 Acknowledgements Thanks to Pratt & Whitney Engines TAMU Turbomachinery Research Consortium Questions (?) Learn more at 30