Annular pressure (damper) seals Stiffness principle. Seals as load support elements. Rotordynamic effects.

Similar documents
Failure of a Test Rig Operating with Pressurized Gas Bearings: a Lesson on Humility

Introduction to rotordynamics and lubricated elements

A Test Rig for Evaluation of Thrust Bearings and Face Seals

Notes 11. High Pressure Floating Ring Oil Seals

Introduction to triblogy, rotordynamics & lubricated elements

EFFECT OF LUBRICANT SUPPLY PRESSURE ON SFD PERFORMANCE: ENDS SEALED WITH O-RINGS & PISTON RINGS

0 INTRODUCTION TO FLUID FILM BEARINGS AND SEALS

Metal Mesh Foil Gas Bearings for Oil-Free Turbomachinery

Structural and Rotordynamic Force Coefficients of a Shimmed Bump Foil Bearing: an Assessment of a Simple Engineering Practice

XLTRC 2 TURBOMACHINERY RESEARCH CONSORTIUM ROTORDYNAMICS SOFTWARE SUITE

COMPARISON OF THE LEAKAGE CHARACTERISTICS OF THE STRAIGHT ANNULAR AND CONVERGENT SEALS. A Thesis SERAFETTIN USTUN

APPLICATION OF A NEW TYPE OF AERODYNAMIC TILTING PAD JOURNAL BEARING IN POWER GYROSCOPE

MEASUREMENTS OF ROTORDYNAMIC ROTOR SUPPORTED ON METAL MESH

ROTATING MACHINERY DYNAMICS

Year I. TRC-SEAL Luis San Andrés. Weilian Shan Graduate Research Assistant. May rd Turbomachinery Research Consortium Meeting

HYDROGEN COMPRESSOR SEAL CASE STUDY UTILIZING HALO (NON-CONTACTING, COMPLIANT) INTER-STAGE, IMPELLER EYE, BUFFER AND FAIL- SAFE SEALS

NOVEL CARBON-GRAPHITE GAS BEARINGS FOR TURBOMACHINERY

Experimental Response of a Rotor Supported on Flexure Pivot Hydrostatic Pad Gas Bearings

Dynamic Coefficients in Hydrodynamic Bearing Analysis Steven Pasternak C.O. Engineering Sleeve and Sleevoil Bearings 8/10/18 WP0281

CHAPTER 1. Introduction and Literature Review

REVAMPING AND PRELIMINARY OPERATION OF A THRUST BEARING TEST RIG

APPLICATION OF A NEW TYPE OF AERODYNAMIC TILTING PAD JOURNAL BEARING IN POWER GYROSCOPE

CONTENTS. 5 BALANCING OF MACHINERY Scope Introduction Balancing Machines Balancing Procedures

EFFECT OFSHIMMING ON THE ROTORDYNAMIC FORCE COEFFICIENTS OF A BUMP TYPE FOIL BEARING TRC-B&C

TRANSLATION (OR LINEAR)

MEASUREMENTS VERSUS PREDICTIONS FOR A HYBRID (HYDROSTATIC PLUS HYDRODYNAMIC) THRUST BEARING FOR A RANGE OF ORIFICE DIAMETERS.

FLUID FLOW. Introduction

Experimental research on dynamic characteristics of gas bearing-rotor with different radial clearances

Applied Fluid Mechanics

ECH 4224L Unit Operations Lab I Fluid Flow FLUID FLOW. Introduction. General Description

A Different Perspective of Synchronous Thermal Instability of Rotating Equipment (STIR) Yve Zhao Staff Machinery Engineer 3/15/2017

(12) Patent Application Publication (10) Pub. No.: US 2012/ A1. Underbakke et al. (43) Pub. Date: Jun. 28, 2012

Rotor Dynamics as a Tool for Solving Vibration Problems Malcolm E. Leader, P.E. Applied Machinery Dynamics Company

Externally Pressurized Bearings and Machinery Diagnostics

SOME INTERESTING ESTING FEATURES OF TURBOCHARGER ROTOR DYNAMICS

Investigations of Oil Free Support Systems to Improve the Reliability of ORC Hermetic High Speed Turbomachinery

Effect of Stator Shape on the Performance of Torque Converter

Study on Flow Fields in Variable Area Nozzles for Radial Turbines

PNEUMATIC HIGH SPEED SPINDLE WITH AIR BEARINGS

Temperature Field in Torque Converter Clutch

Scroll Compressor Oil Pump Analysis

2009. Research on characteristics of fluid-induced vibration for short labyrinth seals

Test Results for Load-On-Pad and Load-Between- Pad Hybrid Flexure Pivot Tilting Pad Gas Bearings

TURBOLINK BEARINGS L U I D I L M E A R I N G S. Manufcturing Engineering Consulting Service & Retrofit

Impacts of Short Tube Orifice Flow and Geometrical Parameters on Flow Discharge Coefficient Characteristics

MARINE FOUR-STROKE DIESEL ENGINE CRANKSHAFT MAIN BEARING OIL FILM LUBRICATION CHARACTERISTIC ANALYSIS

FEASIBILITY STYDY OF CHAIN DRIVE IN WATER HYDRAULIC ROTARY JOINT

Regimes of Fluid Film Lubrication

Stability Analysis of a Turbocharger for Marine Diesel Engine Service. Master of Science in Mechanical Engineering

Design and Test of Transonic Compressor Rotor with Tandem Cascade

ENHANCED ROTORDYNAMICS FOR HIGH POWER CRYOGENIC TURBINE GENERATORS

Technology Readiness of 5 th and 6 th Generation Compliant Foil Bearing for 10 MWE S CO 2 Turbomachinery Systems

May 2015 IDENTIFICATION OF STRUCTURAL STIFFNESS AND MATERIAL LOSS FACTOR IN A LARGE DIAMETER METAL MESH FOIL BEARING. Luis San Andrés and Travis Cable

R10 Set No: 1 ''' ' '' '' '' Code No: R31033

Civil Engineering Hydraulics. Radial Flow Devices

Lecture 4. Lab this week: Review: Pilot-Open-Check. Cartridge valves Flow divider Properties of Hydraulic Fluids. Course feedback (2mins)

Evaluating and Correcting Subsynchronous Vibration in Vertical Pumps

Lecture 4. Lab 8 Check valve and pilot-operated check valves Lab 9 Flow divider. Update: Identifying lab objectives Review: Metering/Bleed-off

November 8, 2018 GAS TURBINE ENGINE SECONDARY FLOW SYSTEMS

Xueliang Lu Research Assistant. Luis San Andrés Mast-Childs Chair Professor Principal Investigator

Modeling tire vibrations in ABS-braking

Pump Control Ball Valve for Energy Savings

Lecture 6. Systems review exercise To be posted this weekend Due next Friday (3/6)

Familiarize yourself with the pressure loss phenomenon. The Discussion of this exercise covers the following point:

Magnetic Bearings for Supercritical CO2 Turbomachinery

III B.Tech I Semester Supplementary Examinations, May/June

Design of A New Non-Contact Screw Seal and Determination of Performance Characteristics

LECTURE-23: Basic concept of Hydro-Static Transmission (HST) Systems

Improvements for Ver November 23, 2017

T E C H N I C A L P A P E R

APPLICATION OF STAR-CCM+ TO TURBOCHARGER MODELING AT BORGWARNER TURBO SYSTEMS

Modern Approach to Liquid Rocket Engine Development for Microsatellite Launchers

FRONTAL OFF SET COLLISION

Seals Stretch Running Friction Friction Break-Out Friction. Build With The Best!

Lecture 6. This week: Lab 13: Hydraulic Power Steering [ Lab 14: Integrated Lab (Hydraulic test bench) ]

Numerical Simulation and Performance Analysis of Rotary Vane Compressors for Automobile Air Conditioner

Turcon Glyd Ring. Double Acting Rubber Energized Plastic Faced Seal Material: Turcon, Zurcon and Elastomer

DYNAMIC ANALYSIS OF A TURBOCHARGER IN FLOATING BUSHING BEARINGS

Test Which component has the highest Energy Density? A. Accumulator. B. Battery. C. Capacitor. D. Spring.

Fluid Power System Model-Based Design. Energy Efficiency. Fluid Power System Model-Based Design Energy Efficiency. K. Craig 1

EFFECT OF SURFACE ROUGHNESS ON PERFORMANCE OF WIND TURBINE

COMPRESSIBLE FLOW ANALYSIS IN A CLUTCH PISTON CHAMBER

Development of Rattle Noise Analysis Technology for Column Type Electric Power Steering Systems

LESSON Transmission of Power Introduction

Constructive Influences of the Energy Recovery System in the Vehicle Dampers

Tilting Pad Journal Bearings

COMPUTATIONAL FLOW MODEL OF WESTFALL'S 2900 MIXER TO BE USED BY CNRL FOR BITUMEN VISCOSITY CONTROL Report R0. By Kimbal A.

CLOSED CIRCUIT HYDROSTATIC TRANSMISSION

Ball Rail Systems RE / The Drive & Control Company

QMOT QSH5718 MANUAL. QSH mm 2.8A, 0.55Nm mm 2.8A, 1.01Nm mm 2.8A, 1.26Nm mm 2.8A, 1.

Experimental Response of a Rotor Supported on Rayleigh Step Gas Bearings

Heat Transfer in Rectangular Duct with Inserts of Triangular Duct Plate Fin Array

FLOW CONTROL THROUGH VORTEX SHEDDING INTERACTION OF ONE CYLINDER DOWNSTREAM OF ANOTHER. Jonathan Payton 1, and *Sam M Dakka 2

Reduction of Self Induced Vibration in Rotary Stirling Cycle Coolers

Heat Transfer Enhancement for Double Pipe Heat Exchanger Using Twisted Wire Brush Inserts

B.TECH III Year I Semester (R09) Regular & Supplementary Examinations November 2012 DYNAMICS OF MACHINERY

Optimization of Packed Tower Inlet Design by CFD Analysis. Dana Laird Koch-Glitsch, Inc.

Balancing and over-speed testing of flexible rotors

QMOT STEPPER MOTORS MOTORS

May 2017 Year III A TEST RIG FOR EVALUATION OF FOIL THRUST BEARINGS: DYNAMIC STIFFNESS OF A METAL MESH THRUST FOIL BEARING TRC-B&C

Transcription:

Modern Lubrication Theory FALL 014 TAMU Date: Today s material Notes 1a Annular pressure (damper) seals Stiffness principle. Seals as load support elements. Rotordynamic effects. Notes 1b Hydrostatic / hydrodynamic bearings Stiffness principle. Effect of fluid compressibility. Force coefficients Observations/Announcements Homework 4, Homework 5 due Read: Chupp et al., 006, AIAA JPP, Childs, Chp 4, pp. 7-84 1

Design and Analysis of High Speed Pumps Damper Seals and Hydrostatic Bearings for Pump Applications Dr. Luis San Andres Mast-Childs Professor Presentation for lectures 1(a) and 1(b) Based on Lecture (3) delivered at Von Karman Institute (006) (Nov update)

Damper Seals & Hydrostatic Bearings LEARNING OBJECTIVES a) Physical mechanism for generation of direct stiffness in annular pressure seals & hydrostatic bearings. Select design conditions to obtain maximum (optimum) stiffness b) Bulk-flow equations for prediction of the flow and force coefficients in annular pressure seals and hydrostatic bearings c) Predictions for two water seals, long and short, for application as neck ring and interstage seals. Effect of seal length and inlet swirl on rotordynamic force coefficients d) Discussion on design of hydrostatic bearing for water pump and to replace oil-lubricated bearings. Effect of angled injection and considerations for improvements in stability margin 3

Annular pressure seals Seals (annular, labyrinth or honeycomb) separate regions of high pressure and low pressure and their principal function is to minimize the leakage (secondary flow); thus improving the overall efficiency of a rotating machine extracting or delivering power to a fluid. Inter-stage seal Impeller eye or neck ring seal Balance piston seal Figure 1 Seals in a Multistage Centrifugal Pump or Compressor 4

Count the seals.. 5

Keep counting seals.. 6

Keep counting more seals.. 7

Keep counting... 8

And keep counting seals.. 9

Annular Pressure Seals Non-contacting fluid seals are leakage control devices minimizing secondary flows in turbomachines. Seals use process liquids of light viscosity as the working fluid. The dynamic force response of pressure seals has a primary influence on the stability response of highperformance turbomachinery. Annular seals, although geometrically similar to plain journal bearings, show a flow structure dominated by turbulence and fluid inertia effects. Operating characteristics unique to seals are * large axial pressure gradients, * large clearance to radius ratio (R/c) < 500, while * the axial development of the circumferential velocity determines the magnitude of crosscoupled (hydrodynamic) forces. 10

Annular Pressure Seals Due to their relative position within a rotor-bearing system, seals modify sensibly the system dynamic behavior. Seals typically "see" large amplitude rotor motions. This is particularly important in back-to-back compressors and long-flexible multiple stage pumps. Figure 1 Straight-Through and Back-to-back Compressors and 1st Mode Shapes 11

Annular pressure seals Intentionally roughened stator surfaces (macro texturing) reduce the impact of undesirable cross-coupled dynamic forces and improve seal stability. These seals are common practice in current damper seal technology for cryogenic turbo pumps. Annular seals acting as Lomakin bearings have potential as support elements (damping bearings) in high speed cryogenic turbo pumps as well in process fluid applications. Figure 3 Honeycomb and round hole pattern seals for turbopump 1

The Lomakin effect in pressurized seals The Lomakin Effect: generation of support (direct) stiffness L Axial velocity, V z shaft c D P s Axial velocity, V z c P a Process P s fluid at high Ps P a P e L Exit pressure P a z 4: Geometry of an annular pressure seal 5: Inertial pressure drop due to a sudden contraction The direct stiffness is due to the pressure drop at the seal inlet plane and its close interaction with the pressure drop (and flow resistance) within the seal lands (even without shaft rotation). The entrance effect is solely due to fluid inertia accelerating the fluid from the upstream (stagnant) pressure supply to a flow with a high axial speed and reduced static pressure. Figures 4 & 5 13

The Lomakin effect: a centering stiffness Force - P supply C V z eccentricity Inlet pressure drop Bernoulli equation: P+ρ (1+)V z /=const. continuity equation: ρv z c=const. + P inlet C V z c V z P - P inlet (1+ξ) ρv z / (1+ξ)ρV z / V z L / C Ω + c V z P Original as per Dr. M. Arghir presentation (006) P exit 14

Centering stiffness from an annular seal K B L c o K P s P a 0.4 0.3 Optimum stiffness * Typical design condition 0. 0.1 pressure drop at inlet Too large clearance Or short seal 0 0 0. 0.4 0.6 0.8 1 pe: pressure ratio (entrance/supply) pressure drop in land Too tight clearance Or Long seal Seal wear enlarges clearance and increases leakage Figure 6 Stiffness versus entrance pressure ratio (simple model) 15

Annular Pressure Seals The Lomakin effect Incompressible fluid, turbulent flow V z = (1 + P s P ) + a L c f z Axial velocity (leakage) decreases with increase of inlet loss () and friction factor (fz) and length of land in seal. P e P a = 1 + P s P a (1+ ) f z c L Entrance pressure into seal decreases as inlet loss () increases and as land friction factor (fz) decreases or land length increases L Ps c Pa Axial velocity, V z Ps Pe 5: Inertial pressure drop due to a sudden contraction Pa z 16

Force Coefficients in Annular Seals D D c P e P s rotor L L stator Depiction of an annular pressure seal P a V z Axial velocity Axial pressure field Y Seal reaction forces are functions of the fluid properties, flow regime, operating conditions and geometry. For small amplitudes of rotor lateral motion: forces represented with linearized stiffness, damping and inertia force coefficients: - F F x y K K xx yx K K xy yy C Y C xx yx C C xy yy M Y M xx yx M M xy yy Y 17

18 Force/Moment coefficients in seals Seals also generate moment coefficients due to tilts (angulations) of rotor. Most complete model: 16 stiffness, 16 damping and 16 inertia coefficients Y Z Y Seal with dynamic translations (,Y) and angulations (, Y) Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y e e M M M M M M M M M M M M M M M M - e e C C C C C C C C C C C C C C C C - e e K K K K - K K K K K K K K K K K K = - M M F F Figure 7

19 Bulk-flow Analysis: governing equations 0 t h hv z hv x z x z V V x V t V h U V h x P h z x x x J x x z V x V V t V h V h z P h z z x z z z Flow Continuity Circumferential Momentum transport Axial momentum transport Pa - Turbulent flow with fluid inertia effects - Mean flow velocities average across film (h) - No accounting for strong recirculation zones z Vz Ps Vx

Bulk-flow analysis: inlet conditions Boundary Conditions -Inlet pressure loss due to fluid inertia (Lomakin effect) - Inlet swirl determined by upstream condition (implementation of swirl-brakes) -Exit pressure without recovery loss, typically P e P s - 1 (1+ ) V z, V x R Numerical Solution -Analytical solutions for short length seals (centered condition) See Childs textbook -Numerical solutions for realistic geometries use CFD techniques (staggered grids, upwinding, etc) and predict (4 or 16) K,C,M force/moment coefficients. Radial baffles retarding fluid swirl Fluid path Ps Seal z Vz Vx Rotor speed rotor Figure 8 Anti swirl brake at inlet or pressure seal 0

Dynamic forced performance of annular seals Example: (centered seal) Fluid: water at 30 C ( 0.79 cpoise, 995 kg/m3) Centered seal (e=0): No static load D = 15. 4 mm (6 inch), Short seal (neck ring seal) L/D=0.0 Long seal (inter stage seal) L/D= 0.50 c = 0.190 mm, nominal clearance & worn clearance (c) smooth rotor and stator surfaces Nominal speed = 3600 rpm, Pressure drop 34.4 bar neck ring seal & inter stage seal or balance piston Standard design practice: find influence of seal wear on leakage and rotordynamic coefficients Pressure drop P ~ RPM Inlet loss coefficient ξ=0.1 Inlet swirl α=0.5 and 0.0 (without and with swirl break) Influence on cross-coupled stiffness and stability Table 1 Geometry and operating conditions of water seals in a liquid pump 1

Entrance pressure into seal LONG SEAL SHORT SEAL entrance pressure 70 60 50 40 30 0 10 Pressures vs shaft speed bar operating speed 3600 rpm 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) water seal, L/D=0.50, c=0.190 mm, D=15 mm No effect of inlet swirl on entrance pressure c c RPM Ps Supply pressure Supply pressure Nominal clearance (C) Twice clearance (worn) L/D=0.50 entrance pressure 70 60 50 40 30 0 10 Pressures vs shaft speed bar 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) water seal, L/D=0.0, c=0.190 mm, D=15 mm No effect of inlet swirl on entrance pressure operating speed 3600 rpm c Supply pressure Supply pressure Nominal clearance (C) Twice clearance (worn) - Worn seal leads to lower entrance pressure loss due to increase in flow rate (reduction in land flow resistance) - Short seal leads has lowest entrance pressure due to increase in leakage c L/D=0.0 Figure 9 Supply and entrance pressures for two water seals, L/D=0.50 and 0.0, and two clearances (c and c)

Entrance pressure ratio into seal Entrance pressure ratio LONG SEAL Inlet Pressure vs shaft speed 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0. 0.1 c c (Pe-Pa)/(Ps-Pa) 0.0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) RPM water seal, L/D=0.50, c=0.190 mm, D=15 mm No effect of inlet swirl on entrance pressure operating speed 3600 rpm nominal C, swirl=0.5 nominal C, swirl=0.0 Twice clearance (worn) swirl=0.5 L/D=0.50 Entrance pressure ratio Inlet Pressure vs shaft speed 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0. 0.1 SHORT SEAL c c (Pe-Pa)/(Ps-Pa) 0.0 0 1000 000 3000 4000 5000 6000 L/D=0.0 rotor speed (RPM) water seal, L/D=0.50, c=0.190 mm, D=15 mm No effect of inlet swirl on entrance pressure operating speed 3600 rpm nominal C, swirl=0.5 nominal C, swirl=0.0 Twice clearance (worn) swirl=0.5 RPM - Worn seal leads to higher entrance pressure loss due to increase in flow rate (reduction in land flow resistance) - Short seal leads to even larger inlet pressure loss due to increase in leakage Figure 9 Entrance pressure ratio for two water seals and two clearances (c and c) 3

Seal leakage LONG SEAL SHORT SEAL 18 16 Leakage vs shaft speed kg/s Nominal condition operating speed 3600 rpm water seal, L/D=0.50, c=0.190 mm, D=15 mm No effect of inlet swirl Nominal clearance (C) Twice clearance (worn) 18 16 Leakage vs shaft speed kg/s water seal, L/D=0.0, c=0.190 mm, D=15 mm No effect of inlet swirl operating speed 3600 rpm 14 14 Flow rate 1 10 8 6 4 c c Flow rate 1 10 8 6 4 c c Nominal clearance (C) Twice clearance (worn) 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) L/D=0.50 0 0 1000 000 3000 4000 5000 6000 L/D=0.0 rotor speed (RPM) RPM & Psupply - Leakage is proportional to RPM ~ P & not proportional to clearance - Worn seal leaks more Longer seal leaks less - No influence of inlet swirl 1/ Figure 10 Leakage (flow rate) for two water seals, L/D=0.50 & 0.0, and two clearances 4

Drag power LONG SEAL SHORT SEAL Drag Power 6 5 4 3 Power vs shaft speed kw Nominal condition c c water seal, L/D=0.50, c=0.190 mm, D=15 mm minor effect of inlet swirl operating speed 3600 rpm Drag Power 6 5 4 3 Power vs shaft speed kw water seal, L/D=0.0, c=0.190 mm, D=15 mm minor effect of inlet swirl operating speed 3600 rpm Nominal clearance (C) Twice clearance (worn) "" no swirl 1 Nominal clearance (C) Twice clearance (worn) "" no swirl 1 c c 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) L/D=0.50 0 0 1000 000 3000 4000 5000 6000 L/D=0.0 rotor speed (RPM) RPM & Psupply - Drag Power ~ RPM - Long seal draws more power (~ 5 times) - No effect of clearance. As C increases, so does circ. Reynolds # - No effect of inlet swirl Figure 11 Drag Power for two water seals, L/D=0.50 & 0.0, and two clearances 5

Rotordynamic coefficients lateral motions Seal reaction forces: - F F x y K K xx yx K K xy yy C Y C xx yx C C xy yy M Y M xx yx M M xy yy Y Reduced model: K = KYY, KY = -KY C = CYY, CY = -CY M = MYY, MY = -MY Whirl frequency ratio WFR ~ Centered Condition KY C Assumes: No static load capability Circular centered orbit : measure of rotordynamic stability Y 6

Seal direct stiffness LONG SEAL SHORT SEAL Direct Stiffness vs shaft speed 60 50 Kxx=Kyy [MN/m] Nominal condition water seal, L/D=0.50, c=0.190 mm, D=15 mm negligible effect of inlet swirl operating speed 3600 rpm Direct Stiffness vs shaft speed 60 50 Kxx=Kyy [MN/m] water seal, L/D=0.0, c=0.190 mm, D=15 mm negligible effect of inlet swirl operating speed 3600 rpm Stiffness 40 30 0 c c Stiffness 40 30 0 Nominal clearance (C) Twice clearance (worn) ""no swirl c c 10 Nominal clearance (C) Twice clearance (worn) ""no swirl 10 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) L/D=0.50 0 0 1000 000 3000 4000 5000 6000 L/D=0.0 rotor speed (RPM) RPM & Psupply - Direct stiffness ~ Pressure supply ~ RPM - Long seal has ~ x larger stiffness than short seal - Worn clearance causes ~ 50% drop in direct stiffness. It will affect pump WET natural frequencies - No effect of inlet swirl 1/ Figure 1 Direct stiffnesses K=KYY for two water L/D seals & two clearances 7

Seal direct stiffness Direct stiffness Dim Direct Stiffness vs pressure ratio 0.50 0.45 0.40 0.35 0.30 0.5 0.0 0.15 0.10 0.05 0.00 Kxx=Kyy [-] L/D=0.5 L/D=0. L/D=0., c c L/D=0.5, c, a=0.5 L/D=0.5, c, a=0.0 L/D=0.5, c, a=0.5 L/D=0.5, c, a=0.0 L/D=0., c, a=0.5 L/D=0., c, a=0.0 L/D=0., c, a=0.5 L/D=0., c, a=0.0 Simple formula water seal, L/D=0.50, 0.0, c=0.190 mm, D=15 mm negligible effect of inlet swirl K c k L D ( P s Pa ) L D ( P c s Pa L/D=0.5, c L/D=0.5, c L/D=0., c 0 0. 0.4 0.6 0.8 1 Entrance Pressure ratio L/D=0.50, 0.0 - Direct stiffness follows simple formula K c ) RPM increases Simple formula Figure Dimensionless stiffnesses K=KYY for two water L/D seals & two clearances 8

Seal cross-coupled stiffness LONG SEAL SHORT SEAL Stiffness Cross Stiffness vs shaft speed 50 45 40 35 30 5 0 15 10 Kxy=-Kyx [MN/m] Nominal condition Nominal clearance (with swirl) "" no swirl Twice clearance (worn) "" no swirl c water seal, L/D=0.50, c=0.190 mm, D=15 mm large effect of inlet swirl operating speed 3600 rpm c 1xC 50% inlet swirl xc no inlet swirl Stiffness Cross Stiffness vs shaft speed 10 8 6 4 Different scale Inlet swirl =0 Kxy=-Kyx [MN/m] Nominal clearance (with swirl) "" no swirl Twice clearance (worn) "" no swirl water seal, L/D=0.0, c=0.190 mm, D=15 mm large effect of inlet swirl operating speed 3600 rpm c c 1xC 50% inlet swirl xc no inlet swirl 5 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) L/D=0.50 RPM & Psupply 0-0 1000 000 3000 4000 5000 6000 rotor speed (RPM) L/D=0.0 - Cross stiffness ~ RPM & 1/clearance - Long seal has ~ 5x more cross-stiffness than short seal - Worn clearance drop in cross-stiffness. - Inlet swirl ~ 0 has most pronounced effect (KY <0 favors stability) Figure 13 Cross stiffnesses KY=-KY for two water L/D seals and two clearances 9

Seal direct damping coefficients LONG SEAL SHORT SEAL Damping 00 180 160 140 10 100 80 60 Damping vs shaft speed Cxx=Cyy Cxy=-Cyx [kns/m] Nominal condition Cxx Nominal clearance (with swirl) Cxy Cxx Twice clearance (worn) Cxy c water seal, L/D=0.50, c=0.190 mm, D=15 mm little effect of inlet swirl c operating speed 3600 rpm Cxx 50% inlet swirl 1xC xc Damping 50 45 40 35 30 5 0 15 Different Damping scale vs shaft speed Cxx=Cyy Cxy=-Cyx [kns/m] Cxx Nominal clearance (with swirl) Cxy Cxx Twice clearance (worn) Cxy water seal, L/D=0.0, c=0.190 mm, D=15 mm little effect of inlet swirl c operating speed 3600 rpm c 1xC 50% inlet swirl Cxx xc 40 0 Cxy 10 5 Cxy 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) RPM & Psupply L/D=0.50 0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) - Direct damping ~ effective turbulent flow viscosity & 1/clearance - Long seal has ~ 5x more direct damping than short seal - Worn clearance drop in damping - Inlet swirl no effect. C >> CY RPM L/D=0.0 Figure 14 Damping C=CYY & CY=-CY for two L/D water seals and two clearances 30

Seal added mass or fluid inertia coefficients Added mass LONG SEAL 50 45 40 35 30 5 0 15 10 5 Direct Inertia vs shaft speed c c Mxx=Myy [kg] 0 0 1000 000 3000 4000 5000 6000 L/D=0.50 Rotor Speed (RPM) water seal, L/D=0.50, c=0.190 mm, D=15 mm negligible effect of inlet swirl operating speed 3600 rpm Nominal clearance (C) Twice clearance (worn) ""no swirl Added mass SHORT SEAL Different Direct Inertia scale vs shaft speed 5.0 4.5 4.0 3.5 3.0.5.0 1.5 1.0 0.5 Mxx=Myy [kg] c c Nominal condition 0.0 0 1000 000 3000 4000 5000 6000 L/D=0.0 rotor speed (RPM) water seal, L/D=0.0, c=0.190 mm, D=15 mm negligible effect of inlet swirl operating speed 3600 rpm Nominal clearance (C) Twice clearance (worn) ""no swirl RPM & Psupply - Direct inertia invariant with speed & proportional to 1/clearance - Long seal has ~ 0x more added mass than short seal - Worn clearance drop in added mass - Inlet swirl no effect. M >> MY - Large inertia will affect wet pump critical speeds Figure 15 Inertia M=MYY for two L/D water seals and two clearances 31

Fluid inertia Its magnitude M fluid DL c M steel steel M D 3 L 1 c tanh L D L D D L D=15.4 mm c =0.190 mm =0.79 cpoise =995 kg/m3 LONG SEAL L/D=0.5 SHORT SEAL L/D=0. 0.5 M 4.03kg M fluid 6.9 10 3 kg M steel 10.84kg M 3.88 M steel 0. M.91kg M fluid.76 10 3 kg M steel 4.34kg M 0.67 M steel -Fluid mass inside film is just a few grams, but. since added mass is proportional to Diameter and 1/clearance M can be larger than mass of solid steel of same dimensions 3

Whirl frequency ratio stability indicator LONG SEAL SHORT SEAL Whirl frequency ratio 0.6 0.5 0.4 0.3 0. 0.1 Whirl ratio vs shaft speed WFR c c Nominal clearance (with swirl) "" no swirl Twice clearance (worn) "" no swirl c c 0.0 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) water seal, L/D=0.50, c=0.190 mm, D=15 mm large effect of inlet swirl operating speed 3600 rpm 0.50 50% inlet swirl 1xC 0. no inlet swirlinlet swirl =0 xc L/D=0.50 Whirl frequency ratio 0.6 0.5 0.4 0.3 0.1 Whirl ratio vs shaft speed WFR Nominal condition Nominal clearance (with swirl) "" no swirl Twice clearance (worn) "" no swirl negative values 0.0 0 1000 000 3000 4000 5000 6000 L/D=0.0 rotor speed (RPM) water seal, L/D=0.0, c=0.190 mm, D=15 mm large effect of inlet swirl operating speed 3600 rpm 50% inlet swirl no inlet swirl xc 1xC 0.50 RPM & Psupply - WFR always 0.50 for inlet swirl = 0.50 Stable up to x critical speed - Inlet swirl = 0.0 drops WFR, in particular for short seal. - Inlet swirl =0.0 does not help greatly in long seal with tight clearance Figure 16 Whirl frequency ratio for two L/D water seals & two clearances 33

Textured surface annular seals Hole-Pattern Seal Unwrap Honeycomb Seal Unwrap -Machined roughness in seal surfaces aids to increase friction thus reducing leakage. - Surface texturing also reduces development of mean circumferential flow velocity, thus decreasing cross-stiffnesses. -Proven improvement in stability margins in pumps and compressors -Texturing works only on stationary surface. Opposite rotordynamic effect if rotor is rough. 34

Hole pattern seal for compressors Hole damper seal replacing labyrinth seal in division wall of back-to-back compressor 005 Turbomachinery Symposium 35

Hole pattern seal INSTABILITY Waterfalls before (with original labyrinth seals) and after installation of holepattern seal in division wall - 36

Questions? 37

Hydrostatic Bearings for pump applications 38

Hydrostatic Bearings for pump applications External pressure source forces fluid to flow between two surfaces, thus enabling their separation and the ability to support a load without contact. Advantages Support very large loads. The load support is a function of the pressure drop across the bearing and the area of fluid pressure action. Load does not depend on film thickness or lubricant viscosity Long life (infinite in theory) without wear of surfaces Provide stiffness and damping coefficients of very large magnitude. Excellent for exact positioning and control. Disadvantages Require ancillary equipment. Larger installation and maintenance costs. Need of fluid filtration equipment. Loss of performance with fluid contamination. High power consumption: pumping losses. Limited LOAD CAPACITY ~ f(psupply) Potential to induce hydrodynamic instability in hybrid mode operation. Potential to show pneumatic hammer instability with compressible fluids 39

Hydrostatic bearings for turbopumps Low cost primary power cryogenic turbo-pumps (TP) are compact, operate at high speeds, and require of externally pressurized fluid film bearings to support radial and thrust loads. Hybrid thrust & radial bearings enable smaller and lighter turbopumps with no DN life limitations Large stiffness (accuracy of positioning) and damping force coefficients allow for unshrouded impellers with increased TP efficiency 40

Support stiffness in a Hydrostatic Bearing P s Q r Feed restrictor Flow through restrictor = Flow through film lands P a L Film land P R, recess pressure b L Fluid flow, Ql Figure 18: Geometry of a simplified 1-D hydrostatic bearing P a h p h Q Q r r Q Q c o A C o d d 4 18 c P P s s P P R R orifice capillar P s Supply pressure P R, recess pressure Q 3 B h P B h PR P 1 x 1 L 3 a Land flow P a Film land L Figure 19: Typical pressure drop in a hydrostatic bearing (laminar flow without fluid inertia effects, incompressible fluid) Figures 18 & 19 b L P a As film thickness decreases, resistance in land increases, thus reducing flow rate and increasing pressure in pocket or recess. Stiffness generated from changes in pocket pressure 41

Static support stiffness in a hydrostatic bearing 0.4 Dimensionless stiffness for simple HB Typical design condition K c Ps Area stiffness (dimensionless) 0. orifice capillar Figure 0: Static stiffness for simple hydrostatic bearing (laminar flow w/o fluid inertia effects, incompressible fluid). Figure 0 0 0 0. 0.4 0.6 0.8 1 ifi pocket pressur ratio (Pr/Ps) Low recess pressure Too large clearance Or too small orifice diameter High recess pressure drop Too tight clearance Or too large orifice No need of journal rotation Stiffness is proportional to supply pressure and bearing (pocket) area. Stiffness is inversely proportional to clearance Stiffness changes quickly with variations in pocket pressure Hydrostatic bearings have LIMIT load capacity 4

Traditional hydrostatic bearing design: * large pocket area (80-90 % of total area) * deep pocket depth * large orifice discharge volume Applications: low or null surface speed, low frequencies, nearly incompressible fluids (water or mineral oil) produces very large DIRECT Stiffness. Warning: This design should NEVER be used with compressible liquids or gases 43

Fluid Compressibility Effects Pneumatic Hammer is a self-excited instability (loss of damping) in poorly designed hydrostatic bearings for applications with compressible liquid and gases. The instability arises due to trapped fluid volume in the bearing pockets generating a dynamic pressure that lags journal motions by ~ (+)90 degrees. Remedies to AVOID (reduce and even eliminate) pneumatic hammer are WELL KNOWN (documented analysis and operation verification) and can be implemented easily at the design stage. 44

Fluid compressibility Effects Low frequency Loss of damping if fluid is compressible Dimensionless stiffness and damping 10 1 0.1 0 1 3 4 K/Ko C/Co Coefficients for hydrostatic bearing C K excitation frequency/break frequency K f ( ) K B 0 f 1 ; 1 f High frequency Stiffness Hardening effect Complete loss of damping C ( ) C 0 1 1 f Coefficients depend on BREAK frequency: B Z 1 V rec 0 3 h0 B 6 L A function of fluid bulk modulus pocket volume Ko C B o Figure Frequency dependent force coefficients for simple hydrostatic bearing 45

Whirl frequency ratio of centered HJB WFR K C Y f K Y 0.5 C 1 0 0 1 1 Whirl frequency ratio Ko C B o B Q P r R 0 Z 1 V rec 0 Z 0 1 V rec 0 3 h0 B 6 L WFR in a hydrostatic bearing can be WORSE than that of a plain journal bearing due to fluid compressibility. Worse conditions are for gases and liquid hydrogen in a cryogenic turbo pump. In aerostatic bearings pockets are not machined to reduce (eliminate) pneumatic hammer For LH, very shallow pockets and reduced pocket area are recommended. 46

Hydrostatic Bearings for Cryogenic Turbo Pumps Radial hydrostatic bearings Thrust hydrostatic bearing Figure 3 Advanced Liquid Hydrogen Turbopump [] 47

Hydrostatic Bearings for Cryogenic Turbo Pumps Radial hydrostatic bearing Thrust hydrostatic bearing Radial hydrostatic bearing for LH TP - Knurled surface Figure 4 Hydrostatic radial and thrust bearings for cryogenic turbopump 48

49 Bulk-flow Analysis of Hydrostatic Bearings 0 t h hv z hv x z x z V V x V t V h U V h x P h z x x x J x x z V x V V t V h V h z P h z z x z z z Flow Continuity Circumferential Momentum transport Axial momentum transport -Turbulent flow with fluid inertia effects -Mean flow velocities average across film (h) - No accounting for strong recirculation zones

Bulk-flow Analysis of Hydrostatic Bearings MR Ps Supply pressure Pocket pressure field with angled injection P s Feed orifice and Supply volume PR, recess pressure HR PR R P R P Film land Fluid flow, M Film land Pa b R H b Figure 6: Turbulent flow pressure distribution in a pocket of a hybrid bearing Flow continuity M R C d A o P P M V s R 1/ t R Pressure rise before edge P R P R x b x h h R R R V Pressure rise at edge P R P R (1 ) e 1 e h h h R V x, z Figure 6 Pocket pressures: angled & radial injection 50

Bulk-flow Analysis of Hydrostatic Bearings Small amplitude radial motions (,Y) at frequency () to derive zeroth order equations for equilibrium flow field: flow rate, load capacity, power loss, temperature raise first-order equations for perturbed flow field: (stiffness, damping and inertia) force coefficients Z P H R dz d K M i C ;,, Y A B Numerical method of solution: SIMPLEC method, extension of control volume method, + component method for superposition of hydrostatic and hydrodynamic effects. Sensitivity Analysis Bearing force coefficients and performance parameters are most sensitive to orifice entrance loss coefficient. Numerical Analysis : Hydrostatic Bearings 51

Dynamic Performance of Hydrostatic Bearings Example: Water bearing replacing oil lubricated bearing Fluid: water at 30 C ( 0.79 cpoise, 995 kg/m3) D=L = 15. 4 mm (6 inch) c=0.10 mm (4 mil), nominal clearance 5 pockets: l= 51 mm, arc 41, depth= 0.381 mm Orifice diameter: 3. mm (Cd=0.80) Static load =Wx = 5000 N smooth rotor and stator surfaces Nominal speed = 3600 rpm, Pressure drop 34.4 bar c = 1.33x oil bearing Pocket depth/c = 3.75 Pocket area = 19% to avoid hammer STATIC LOAD FROM PUMP WEIGHT Pressure supply ~ RPM (bleed off from pump discharge) Table Inlet loss coefficient ξ=0.1 Inlet swirl α=0.50 RADIAL AND TANGENTIAL INJECTION Geometry and operating conditions of HJBs for a liquid pump Influence on cross-coupled stiffness and stability 5

Hydrostatic Bearing Orifice diameter selection Stiffness coefficients HJB design: 400 350 300 50 00 150 100 50 [MN/m] Kxy=-Kyx At nominal operating condition Ps-Pa=34.4 bar, 3600 rpm Design Kxx = Kyy Kxx=Kyy 3.0 mm Selected orifice diameter 0.5 3 3.5 4 4.5 5 5.5 Orifice diameter (mm) L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg) Kxy = -Kyx Orifice diameter Selected diameter Kxx (MN/m) Kxy (MN/m) Stiffness coefficients HJB design: 400 350 300 50 00 150 100 50 Kxx=Kyy= Kyy Kxx (MN/m) Kxy (MN/m) Kxx-Mxx w^ At nominal operating condition Ps-Pa=34.4 bar, 3600 rpm Kxx - Mxx Kxy=-Kyx Kxy = -Kyx 3.0 mm Selected orifice diameter L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg) 0 0. 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Pocket pressure ratio Pocket pressure ratio Orifice diameter - Operating condition: 3.6 krpm & 34.4 bar pressure supply (NO LOAD) - Select orifice diameter to MAIMIZE direct stiffness - No influence of angled injection on K Figure 7 Direct and cross-coupled stiffnesses versus orifice diameter and pocket pressure for water hydrostatic bearing. Nominal operating condition, centered bearing (no load) 53

Hydrostatic Bearing Orifice diameter selection Parameters HJB design: 3.5 3.0.5.0 1.5 1.0 0.5 0.0 Design Pocket pressure ratio HJB flow rate (kg/s) Power (kw) At nominal operating condition Ps-Pa=34.4 bar, 3600 rpm flow selected orifice diameter L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg) Drag power (kw) Flow rate (kg/s) Supply pressure Pocket pressure ratio.5 3 3.5 4 4.5 5 5.5 Orifice diameter (mm) Pocket pressure ratio Parameters HJB design: 1000 900 800 700 600 500 400 300 00 100 0 Damping Mxx = Myy At nominal operating condition Ps-Pa=34.4 bar, 3600 rpm selected design Inertia Mxy = -Myx L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg) Mxx (kg) Mxy (kg) Cxx (kns/m) Cxy (kns/m) Cxx = Cyy Cxy = Cyx.5 3 3.5 4 4.5 5 5.5 Orifice diameter (mm) Pocket pressure ratio Selected diameter - Drag power, flow rate and pocket pressure increase with orifice diameter - Direct damping and inertia coefficients decrease Figure 8 Performance parameters for water hydrostatic bearing versus orifice diameter. Nominal operating condition, centered bearing (no load) 54

Hydrostatic Bearings Static Load Performance Operating journal eccentricity L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm Attitude angle L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm 1.0 0.9 0.8 0.7 operation operating condition, Ps-Pa=34.4 bar 40 35 30 [deg] Radial operating condition, Ps-Pa=34.4 bar 0.6 0.5 0.4 0.3 0. 0.1 0.0 e/c [fraction of clearance] tangential injection radial injection 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn rotor speed (RPM) 5 0 15 10 5 0 Tangent x load direction y tangential injection radial injection 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn rotor speed (RPM) RPM & Psupply RPM & Psupply Load (W)=5 kn - At operating condition, static eccentricity is small, 10% of clearance (c=0.10 mm) - Large eccentricities at low speeds because pressure supply is low (hydrodynamic operation) -Tangential injection reduces attitude angle: will result in improved stability of bearing Figure 9 Journal eccentricity and attitude angle versus rotor speed. Water hydrostatic bearing with radial and tangential injection 55

Hydrostatic Bearings Peak pressures Max film pressures L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm Mass flow rate L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm 80 70 60 [bar] operating condition, Ps-Pa=34.4 operation bar supply pressure function of pump speed 3.0.5 [kg/s] operating condition, Ps- Pa=34.4 bar 50 40 30 0 10 0 Tangent Psupply Radial tangential injection radial injection Supply pressure 0 1000 000 3000 4000 5000 6000.0 1.5 1.0 0.5 0.0 flow R & tang tangential injection radial injection 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn rotor speed (RPM) STATIC LOAD = 5 kn rotor speed (RPM) Load (W)=5 kn - Flow rate is proportional to rotor speed since supply pressure (~ RPM ) - Max. film pressure is proportional to supply pressure, Max. Pressure ~ RPM - Tangential injection increases edge recess pressure no influence on flow rate - Large flow rate (90 LPM) typical of application 1/ Figure 30 Maximum film pressures and flow rate versus rotor speed. Water hydrostatic bearing with radial and tangential injection 56

Hydrostatic Bearings Drag Torque Drag Torque 16 14 1 10 8 6 4 [N-m] operating condition, Ps-Pa=34.4 bar tangential injection radial injection 0 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn operation Tangent L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm Drag Power Radial rotor speed (RPM) 8 7 6 5 4 3 [kw] operating condition, Ps-Pa=34.4 bar L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm Radial Tangent tangential 1 injection radial injection 0 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn rotor speed (RPM) Load (W)=5 kn - Drag torque & power proportional to RPM - Tangential injection decreases power & torque since it retards development of circumferential speed. Care at low speeds and high pressures (reverse journal rotation) Figure 31 Torque and drag power versus rotor speed. Water hydrostatic bearing with radial and tangential injection 57

Hydrostatic Bearings STIFFNESSES Direct stiffnesses Kxx, Kyy L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm Cross stiffnesses Kxy, -Kyx L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm 900 operating condition, Ps-Pa=34.4 bar [MN/m] 800 operation y 700 x 600 Tangent Radial 500 400 300 Kxx: tangential injection 00 radial injection Kyy tangential injection 100 radial injection 0 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn Rotor Speed (RPM) RPM & Psupply 800 700 600 500 400 300 00 100 0 0 1000 000 3000 4000 5000 6000-100 -00 [MN/m] STATIC LOAD = 5 kn x operating condition, Ps-Pa=34.4 bar y Rotor Speed (RPM) Kxy: tangential injection radial injection -Kyx tangential injection Tangent radial injection radial injection Radial tangential injection RPM & Psupply Load (W)=5 kn - At low speeds, bearing operates under hydrodynamic conditions since feed pressure is too low. All coefficients are large since operating eccentricity (e/c) is large - At operating speed, direct stiffness K~KYY & KY=-KY since (e/c)~0 : near centered operation. - Tangential injection reduces greatly cross-coupled stiffness. Figure 3 1 Stiffness coefficients versus rotor speed. Water hydrostatic bearing with radial and tangential injection 58

Hydrostatic Bearings DAMPING Direct damping Cxx, Cyy L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm Cross damping Cxy, -Cyx L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm 3000 500 000 1500 [kns/m] x y operating condition, Ps-Pa=34.4 bar Cxx: tangential injection radial injection Cyy tangential injection radial injection 300 50 Different scale 00 150 [kns/m] operating condition, Ps-Pa=34.4 bar operation Cxy: tangential injection radial injection -Cyx tangential injection radial injection Radial radial injection 1000 500 0 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn Tangent Radial rotor speed (RPM) 100 50 0 x y 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn Tangent tangential injection rotor speed (RPM) Load (W)=5 kn - At low speeds, all coefficients are large since operating eccentricity (e/c) is large (Hydrodynamic operation) - At operating speed, direct damping C~CYY & CY=-CY since (e/c)~0 : near centered operation. - From moderate to high speeds, C > CY - Tangential injection reduces cross-coupled damping small effect on rotordynamics Figure 33 1 Damping coefficients versus rotor speed. Water hydrostatic bearing with radial and tangential injection 59

Hydrostatic Bearings ADDED MASS Direct inertia 00 180 160 140 10 100 80 60 40 0 [kg] x Mxx: tangential injection radial injection Myy tangential injection y Mxx, Myy Tangent radial injection operation L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm operating condition, Ps- Pa=34.4 bar Radial 0 0 1000 000 3000 4000 5000 6000 Cross-inertia 5 Different [kg] scale 0 15 10 5 Mxy: tangential injection radial injection +Myx tangential injection -Mxy radial injection x y Mxy, -Myx Tangent L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm operating condition, Ps-Pa=34.4 bar Radial 0 0 1000 000 3000 4000 5000 6000 STATIC LOAD = 5 kn Rotor Speed (RPM) STATIC LOAD = 5 kn rotor speed (RPM) Load (W)=5 kn - Large fluid inertia effect (~ 160 kg) >> kg = solid steel piece (L,D) - Direct inertia changes little with rotor speed - At operating speed, direct inertia M~MYY & MY=-MY since (e/c)~0 : near centered operation. - For all speeds, M > MY - Tangential injection reduces cross-coupled inertia small effect on rotordynamics Figure 34 1 Fluid inertia coefficients versus rotor speed. Water hydrostatic bearing with radial and tangential injection 60

Whirl Frequency Ratio & Critical mass Whirl frequency ratio L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm Critical mass L=D=0.15 m, c=10 um, 5 pocket (l=l/3, arc =4 deg), orifice do=3. mm 0.70 0.60 [-] Radial 1000000 [kg] operating condition, Ps-Pa=34.4 bar 0.50 0.40 0.30 0.0 Tangent tangential injection radial injection tangential injection against shaft rotation 100000 10000 Tangent tangential injection radial injection 0.10 0.00 Fully stable WFR =<0 operating condition, operation Ps-Pa=34.4 bar 0 1000 000 3000 4000 5000 6000 1000 Radial 0 1000 000 3000 4000 5000 6000 rotor speed (RPM) STATIC LOAD = 5 kn rotor speed (RPM) Load (W)=5 kn - WFR ~ 0.60 for radial injection bearing, Critical mass ~ 8,000 kg - With tangential injection: WFR drops rapidly towards 0.10 at operating point and beyond. Critical mass at least 10x larger than for radial injection - In other applications (ex: LOx bearing), benefit of tangential injection is lost at very high speeds when hydrodynamic effects dominate flow -Critical mass results applicable to RIGID ROTOR only Figure 35 Whirl frequency ratio & Critical Mass versus rotor speed. Water hydrostatic bearing with radial and tangential injection 61

Hydrostatic Bearings Recommendations Hydrostatic bearings have WFR > ~ 0.50 limiting their application to ~x critical speed. Limiting speed condition can be worse if fluid is compressible and pockets are too deep & large area. To reduce risk of hydrodynamic instability & increase bearing stability margin: -Texture bearing surface Proven with macro rough surfaces such as Knurled, round hole and tire truck pattern tested successfully -Angled injection against rotation Retards circumferential flow swirl, effectiveness reduces at high rotor speeds, Can induce backward whirl. Tested successfully -Bearing asymmetry Geometrically induce stiffness orthotropy (K > KYY) Axial feed grooves, mechanical preload, etc. Tested & patented! 6

Hydrostatic Bearings Recommendations Flexure-pivot Tilting pad hybrid bearing Tilting pads accommodate shaft motions to load direction, no generation of cross coupled stiffnesses, KY ~ KY ~ 0. No stability margin. Wire EDM construction allows for reliable hydrostatic feeding. Tested in water, oil and gas applications 63

Hydrostatic Bearings for Cryogenic Turbo Pumps To avoid pneumatic hammer: * reduce area pocket/land area to ~ 15-5% * pocket depth to clearance ratio ~ 10 or less * design and construct orifices without supply discharge volume (no pressure recovery zone) P a L Film land P s Q r P R, recess pressure Feed restrictor L b Fluid flow, Ql P a h p h Design parameter, P < < 1, TYP 0.1 or less DP = (PR -PA) 3 Npocket (Vpocket+Vsupply orifice) Kfluid (Z+1) Area_bearing x Cradial_clearance Kfluid: Fluid Bulk modulus Area_bearing ~ D L Z= (PR -PA) a (PS -PR) a= for orifice restrictors PS, PR, PA: supply, recess or pocket, discharge pressures 64

Hydrostatic Bearings for Cryogenic Turbo Pumps Cryogenic fluid hydrostatic bearing design: * small pocket area (15-5 % of total area) * shallow pocket depth * small or null orifice discharge volume Application: high surface speeds, low and high frequencies, compressible liquids (LO, LH, LN) + Angled injection against rotation to reduce cross-coupled stiffnesses (avoid hydrodynamic instability) Nearly inherent restrictor type, i.e. orifice coefficient regulated by clearance 65

Gas hydrostatic bearing design * Null pocket area (0 % of total area) * Inherent restrictor type, i.e. orifice coefficient regulated by clearance Applications: low and high surface speeds, low and high frequencies, All Gases (air, GO, GH, GN) + angled injection against rotation to reduce cross-coupled stiffnesses 66

Hydrostatic Bearings Model Validation HYDROJET radial hydrostatic bearings Tests at TAMU with water (1000 psi (70 bar) max, 5 krpm max). + 0 bearings x 3 clearances & pocket depths, different pocket shapes, macro-roughness (surface textured) bearings, angled injection. Gas Honeycomb seals Water Lomakin Bearings (Snecma-SEP) Oil tilting and flexure pivot journal bearings HYDROTHRUST axial thrust hydrostatic bearings NONE available in literature for high speed, high pressure (turbulent flows) Concerns: centrifugal and advection fluid inertia cause severe fluid starvation in bearing and reduced axial stiffness coefficients due to effect of added mass coefficients TESTs verified predictions 67

Needs in turbomachinery largest power output to weight ratio, reliability and performance, compact with low number of parts, extreme temperatures and pressures, Rotordynamics low friction and wear automated agile processes coatings: nanopowders surface conditioning environmentally safe and friendly. oil-free machinery inert gas buffer sealing Desired features of support elements (bearings and seals) High stiffness and damping coefficients Linearity with respect to amplitudes of rotor motion Avoidance of rotordynamic instability due to hydrodynamic effects Controllable features to avoid surge/stall, etc. 68

Hydrostatic Bearings Learn more See http://rotorlab.tamu.edu Publications For complete list of computational model predictions and comparisons to test data (over 30 journal papers and 10+ technical progress reports to NASA, USAF, P&W, Rocketdyne, Snecma-SEP, Northrop Grumman) 69

Bulk-flow Analysis of Hydrostatic Bearings At Texas A&M: Hydrojet & Hydrothrust Equations for flow in film lands of bearing Equations for flow in pockets of hydrostatic bearing Flow conditions Mass conservation, Bulk-Flow momentum in circumferential and axial directions (D), Energy transport for mean flow temperature Various surface temperature models Fluid inertia effects at entrance and exit flow regions. Global mass conservation: orifice inlet flow, flow from recess towards or from film lands, and rate of accumulation of fluid within pocket volume, Global momentum in circumferential direction due to angled injection. Global energy transport with adiabatic heat flow surfaces Laminar, laminar to turbulent transition and fully developed turbulent bulk-flow model. Turbulent flow closure model: Moody's friction factor including surface roughness. Fluid with variable properties f(p,t) Model for Cryogenic Hydrostatic Bearings 70

Hydrostatic Bearings Funding & Work to Date 40 k, Rocketdyne (1988-1991), 110 k, Pratt & Whitney (1991-9), 360 k, NASA GRC (1993-1996), 10 k, NASA MSFC (1998/99-001/) 9 k, Norhtop Grumman (005-007) - (USET Program) Separate funding 1988-1995 for performance evaluation and rotordynamics measurements for validation of radial bearing tool prediction All US turbo pump manufacturers and NASA, including SNECMA-SEP, use Hydrojet and Hydrothrust to model cryogenic fluid film bearings and seals. Other industries and Universities have benefited from technology. USET Program (005-008) CLIN 4..1.3. (a) non-linear forced response of fluid film bearing (98.5 k) CLIN 4..1.3. (b) mixed flow regime lift off response (130.5 k) CLIN 4..1.3.7 Experimental Study of Hydrostatic / Hydrodynamic Thrust Bearings (788 k) 71

Hydrostatic Bearings CFD models 3D CFD modeling of flow and pressure in POCKETS carried out by P&W (90 s) and also by Universite de Poitiers funded by Snecma- SEP (France, 001-005). OBJECTIVEs: Predict complex flow field in pocket and extract empirical parameters for pressure loss at inlet to film lands and for development of pressure profile for angled injection bearings. Insert empirical parameters into D-bulk-flow codes for prediction of bearing rotordynamic force coefficients. Best published work: Mihai ARGHIR, Université de Poitiers, France. FINDINGS: 3D-CFD predictions are NOT better than D-bulk flow predictions when compared to test data (TAMU-Childs). TAMU codes are still unsurpassed in terms of accuracy of predictions and speed of execution. 7