I I. Early Shaft Crack Detection On Rotating Machinery Using Vibration Monitoring and Diagnostics _. ) region. acceptance

Similar documents
Vibration based condition monitoring of rotating machinery

Applications Note. Bently Nevada* Asset Condition Monitoring. Periodically Monitored Assets. Connecting SCOUT to Continuous Monitoring Systems

Presented By: Michael Miller RE Mason

3500/46M Hydro Monitor

What are we looking at?

An Introduction to Time Waveform Analysis

Overall vibration, severity levels and crest factor plus

Bently NevadaTM. Technical Training

Beating Phenomenon of Multi-Harmonics Defect Frequencies in a Rolling Element Bearing: Case Study from Water Pumping Station

Getting the Best Performance from Challenging Control Loops

Emerson Process Management - CSI

DETECTION OF ROTOR CRACKS. Donald E. Bently. and Agnes Muszynska. Chief Executive Officer and President. Senior Research Scientist

How Plant Rotating Equipment Resonance Issues Can Affect Reliability and Uptime

Overview of condition monitoring and vibration transducers

PeakVue Analysis for Antifriction Bearing Fault Detection

Acceleration Enveloping Higher Sensitivity, Earlier Detection

Vibration History. Pulp & Bleach Area. ips. Average Amplitude Velocity. Year

Application of Electrical Signature Analysis. Howard W Penrose, Ph.D., CMRP President, SUCCESS by DESIGN

Condition Monitoring of Rotationg Equpiment s using Vibration Signature Analysis- A Review

Industrial vibration sensor selection: Piezovelocity transducers

CASE STUDY: Rotor Bar Fault in AC Induction

Machinery Fault Diagnosis

A Novel Approach to Electrical Signature Analysis

Setpoint Rolling Element Bearing Support

RESEARCH PAPER CONDITION MONITORING OF SIGLE POINT CUTTING TOOL FOR LATHE MACHINE USING FFT ANALYZER

ZOOM Software Measurement and Graph Types

Torsional Monitoring of Turbine-Generators for Incipient Failure Detection. Prepared for:

PROTECTION RELAY FOR SHAFT CURRENT AND VOLTAGE

Bearing fault detection of wind turbine using vibration and SPM

Application Note. Case study Integrated vibration, process monitoring at HPP Momina Klisura

Also, side banding at felt speed with high resolution data acquisition was verified.

What you discover today determines what you do tomorrow! Potential Use of High Frequency Demodulation to Detect Suction Roll Cracks While in Service

Rotordynamics Analysis Overview

Industrial vibration sensor selection: piezovelocity transducers

Practical Machinery Vibration Analysis and Predictive Maintenance

RELIABILITY WEEKLY 2 MACHINE RESONANCE & VIBRATIONS

DEVELOPMENT OF A ROTORDYNAMIC SIGNAL PROCESSING MATLAB INTERFACE AND A TWO-DISK ROTOR MODEL. A Thesis. presented to

Classification of Misalignment and Unbalance Faults Based on Vibration analysis and KNN Classifier

AUTOMATED BEARING WEAR DETECTION. Alan Friedman

AGN 008 Vibration DESCRIPTION. Cummins Generator Technologies manufacture ac generators (alternators) to ensure compliance with BS 5000, Part 3.

Shaft Vibration Monitoring System for Rotating Machinery

CONDITION MONITORING A. R.L Thanmei Sr.E (Telecom), NT

Copyright 2017 by Turbomachinery Laboratory, Texas A&M Engineering Experiment Station

Automated Bearing Wear Detection

SVENSK STANDARD SS :2014

Application Note. Monitoring strategy Diagnosing gearbox damage

Bearing Failures... root causes and preventive procedures Seminar Synopsis

Detection of Wind Turbine Gear Tooth Defects Using Sideband Energy Ratio

Troubleshooting accelerometer installations

Real-time Detection of Developing Cracks in Jet Engine Rotors By: Eric Sonnichsen of Test Devices, Inc. Reprinted from: IEEE Conference 2000

ORBIGate The Turbomachinery Vibration Software

BLADE AND SHAFT CRACK DETECTION USING TORSIONAL VIBRATION MEASUREMENTS PART 2: RESAMPLING TO IMPROVE EFFECTIVE DYNAMIC RANGE

On-line Flux Monitoring of Hydro-generator Rotor Windings

Analysis Of Induction Motor With Broken Rotor Bars Using Discrete Wavelet Transform Princy P 1 and Gayathri Vijayachandran 2

Bearing Fault Diagnosis

Prediction of Defects in Roller Bearings Using Vibration Signal Analysis

EXPERIMENTAL RESULTS OF TURBO-AGGREGATE VIBROACOUSTIC DIAGNOSIS OBTAINED WITH VIBRO-EXPERT SYSTEM FOR ONE TURBO AGGREGATE IN LUKOIL REFINERY

There s Still Value in Overall Vibration Measurements By John C. Johnson Balance Plus Wichita, Kansas

Vibration Analysis Of Rotating Machines With Case Studies

CONTINUOUS CONDITION MONITORING WITH VIBRATION TRANSMITTERS AND PLANT PLCS

Condition Monitoring and Vibrational Analysis of Shaft Through Experimental and FEA Approach

Optical Encoder Applications for Vibration Analysis

Machine Diagnostics in Observer 9 Private Rules

Rotating Machinery Analysis

Predictive Maintenance with Multi-Channel Analysis in Route and Analyze Mode

VIBRATIONAL MEASUREMENT ANALYSIS OF FAULT LATENT ON A GEAR TOOTH

ON TWO-PLANE BALANCING OF SYMMETRIC ROTORS

Low Frequency Velocity Sensor

Of interest in the bearing diagnosis are the occurrence frequency and amplitude of such oscillations.

CHAPTER 4 POWER QUALITY AND VAR COMPENSATION IN DISTRIBUTION SYSTEMS

High Temperature Velocity and Acceleration Sensor

Investigation and Analysis of Chatter Vibration in Centerless Bar Turning Machine

COMPARATIVE PERFORMANCE OF SMART WIRES SMARTVALVE WITH EHV SERIES CAPACITOR: IMPLICATIONS FOR SUB-SYNCHRONOUS RESONANCE (SSR)

Mechanical vibration Rotor balancing. Part 31: Susceptibility and sensitivity of machines to unbalance

CHAPTER 7 FAULT DIAGNOSIS OF CENTRIFUGAL PUMP AND IMPLEMENTATION OF ACTIVELY TUNED DYNAMIC VIBRATION ABSORBER IN PIPING APPLICATION

NON-SELLABLE PRODUCT DATA. Order Analysis Type 7702 for PULSE, the Multi-analyzer System. Uses and Features

Active Vibration Isolation of an Unbalanced Machine Tool Spindle

MECH-303: Gaskets, Packing and Mechanical Seal Failures Analysis

ANALYSIS OF MACHINERY HEALTH CONSIDERING THE PARAMETERS OF VIBRATION IN A MULTI-FUNCTIONING ARRANGEMENT

Enayet B. Halim, Sirish L. Shah and M.A.A. Shoukat Choudhury. Department of Chemical and Materials Engineering University of Alberta

Research Article High Frequency Acceleration Envelope Power Spectrum for Fault Diagnosis on Journal Bearing using DEWESOFT

Artesis Predictive Maintenance Revolution

Appearance of wear particles. Time. Figure 1 Lead times to failure offered by various conventional CM techniques.

of harmonic cancellation algorithms The internal model principle enable precision motion control Dynamic control

Analysis of Power Oscillations and Their Treatment in Hydro Power Plant with Large Bulb Turbines

1. Random nature as to polarity, in that the pulse is sometimes positive and sometimes negative with respect to ground.

Electrical Machines Diagnosis

The Multi-Technology Approach to Motor Diagnostics

CASE STUDY OF OPERATIONAL MODAL ANALYSIS (OMA) OF A LARGE HYDROELECTRIC GENERATOR

Stator Fault Detector for AC Motors Based on the TMS320F243 DSP Controller

PHYSICAL PHENOMENA EXISTING IN THE TURBOGENERATOR DURING FAULTY SYNCHRONIZATION WITH INVERSE PHASE SEQUENCE*

Introduction*to*Machinery*Vibration*Sheet*Answer* Chapter*1:*Vibrations*Sources*and*Uses*

Motors: The Past. is Present. Hunting in the Haystack. Alignment: Fountain of Youth for Bearings. feb Windows to the IR World

APPLICATION NOTE. Detecting Faulty Rolling Element Bearings. Faulty rolling-element bearings can be detected before breakdown.

Unbalance Detection in Flexible Rotor Using Bridge Configured Winding Based Induction Motor

Vibration Fundamentals Training System

Power Plant and Transmission System Protection Coordination Fundamentals

CHAPTER 3 DEFECT IDENTIFICATION OF BEARINGS USING VIBRATION SIGNATURES

NEW DEVELOPMENTS IN FLUX MONITORING FOR TURBINE GENERATORS. M. Sasic, B. A. Lloyd and S.R. Campbell Iris Power LP, Mississauga, Ontario, Canada

Cylindrical rotor inter-turn short-circuit detection

Transcription:

BENTLY(\ NEVADA V TECHNICAL BULLETIN Early Shaft Crack Detection On Rotating Machinery Using Vibration Monitoring and Diagnostics o acceptance region I I 270......_ / 1X amplitude and phase~~...,;v...,;;",e~ct~o,;;""",ilr --~

TECHNICAL BULLETIN Shaft Crack Detection Trends in machine use and design are causing shaft cracks to become increasingly more commonplace than 20 years ago. The current trend among machinery users is to extend the life of rotating machinery. Rather than replace 20- to 3D-year old machinery, many companies are using life assessment and inspection techniques to operate their machinery up to or beyond its original design life. Today's operating practices--particularly on machines in power generation service--also impose severe thermal and mechanical stress on the rotors. One such practice is the runup and rundown of machines two or more times per day. When operated in this manner, some rotors become susceptible to age-related cracking. Another trend over the past 10 years has been to design larger turbine generators--even greater than 1,000 megawatts. These larger machines are more susceptible to various machine malfunctions, including shaft cracks, according to an Electric Power Research Institute (EPRI study. Conseguently, the number of shaft crack incidents has increased dramat i ca lly. One machine manufacturer has logged more than 28 incidents in North America over the past 10 years in the power generation industry alone. And the manufacturer indicates this is a partial list only. Cracks, which develop in new or overhauled rotors because of faults in design, tend to develop early in the life of the rotor after commissioning or rebuilding. These cracks are often caused by increased stress introduced by poor machining techniques. Cracks can also develop in rotors that have run many thousands of hours without failure, because of corrosion damage, severe misalignment preloads, and other factors. These problems can develop on many types of rotating machinery, both large and small. As the number of shaft crack incidents has increased, so, too, has the need for early detection of shaft cracks. The reason is simple. The consequences of shaft cracks are catastrophic. Companies--from a monetary, safety, and public image point of view--cannot afford to have a shaft crack incident. According to an EPRI report, one utility paid $6.2 million to purchase replacement power alone during an outage caused by a shaft crack on a turbine. The cost of instrumentation to diagnose the shaft crack was $100,000. 1

Because other malfunctions can cause the machine to exhibit si~ilar symptoms as those experienced under a shaft crack condition, it's important to understand that proper diagnostic methods must be used to diagnose a shaft crack. There are two fundamental symptoms of a shaft crack: 1 Unexplained changes in the synchronous speed (IX shaft relative amplitude and phase and/or slow roll bow vector, and 2 The occurrence of twice rotative speed (2X vibration. The vital and primarf symptom is changes in the synchronous (IX amplitude and phase and/or s ow roll bow vector. This symptom, which has been observed on large turbine generators with shaft cracks, is the most im ortant indicator of a shaft crack. Shaft measurements are the only effective method for measuring changes in the synchronous IX amp ltude and phase and/or slow roll vector. The changes in synchronous (IX amplitude and phase are caused by the shaft bowi~g due to an asymmetric transverse crack (Figure 1. In this situation, the synchronous (IX amplitude and phase changes- either higher or lower. The secondary, and classical, symptom--the occurrence of the 2X component--is due to asymmetry of the shaft. The 2X component is due to a combination of a transverse crack--which causes shaft asymmetry--and a steady-state radial load--such as gravity, in the case of horizontal rotors. The 2X frequency component is especially dominant when the rotative speed is in the region of half of a rotor system natural frequency. Figure 2 shows a spectrum cascade plot and the corresponding orbits, which identify this classical shaft crack behavior as well as the bow/nonlinear response. The change in synchronous (IX amplitude and phase can be monitored under normal operating conditions to provide alarming and early warning of a shaft crack. The polar plot provides an excellent format for documenting the shift in synchronous (IX amplitude and phase. The shift is monitored at operating speed and at slow roll. A normal operating range of the IX vibration vector is determined within the plot to form what is called an "acceptance region." The actual IX vibration vector is then plotted. Deviation of the IX vibration vector from the acceptance region can be a vital warning of a shaft crack, even though other rotor disturbances can also cause some deviation from the acceptance region (Figure 3. Similar techniques apply to changes in the slow roll bow vector. Changes in synchronous (IX amplitude and phase must be analyzed in conjunction with other vibration information--including the twice rotative speed (2X machine behavior--to determine whether the shifts were caused by an asymmetric transverse crack or other factors, such as load, field current, steam conditions, or other operating parameters. 2

lotted for the twice rotative startu and s utdown to rovide further evi ence ate poss1 i ity a a sat crac. The polar plot format is used to document increases in the twice rotative speed (2X component within an acceptance region. We recommend that the maximum amplitude of the acceptance region for the twice rotative speed (2X component be established at 2 mils peak-to-peak (pp, based on our experience with shaft cracks and other machine malfunctions. When relative changes of the twice rotative speed (2X component exceed 2 pp mils, it is an indication of a severe machine malfunction. It is important to note that the 2 pp mil level may not be appropriate for all machines. The type of machine and the location of the proximity probes on the machine must be considered when establishing the acceptance level. Even though the appearance of the 2X component has been described in technical papers for the past 50 years, only one of four recent shaft crack saves exhibited this classical phenomenon. In that case, the plant engineers and manufacturer's engineer observed and acted upon the following information on a vertical pump used in nuclear power plant service to accomplish an excellent save: 1. Increasing overall vibration levels. 2. Large IX and 2X frequency vibrations. 3. 2X frequency vibration that remained after the machine was trim balanced. The IX frequency vibration was significantly decreased by trim balancing. The engineers used the vibration information to determine whether the machine response was caused by other possible malfunctions, such as unbalance, misalignment, etc. After eliminating the other possible malfunctions, the engineers determined there was a high probability of a shaft crack. The machine was taken out of service and inspected. Inspection confirmed their diagnosis. Observation of orbits is also useful for revealing a shaft crack. Figure 4 shows the typical orbit patterns for a shaft with a large IX component, caused by a crack as well as a steady-state preload, such as gravity. The orbit patterns show that the rotor moves toward the direction of the preload twice during each shaft rotation. This results in a 2X component. Future changes in the orbit pattern are dependent on the phase lag relationship between the IX and 2X components. When these orbit patterns are detected, other machine malfunctions must be eliminated to determine whether a shaft crack is the root cause of the problem. A large Ix component can be caused by a shaft thermal bow as well as a shaft crack. A preload--due to misalignment--as well as a shaft crack can cause large IX and 2X components. 3

When the thermal bow is removed, both IX and 2X components disappear. When only the steady-state load is removed, the orbit becomes a IX circle without a 2X component. The two separate methods of 2X formation may, of course, occur together when both situations are present (see Figure 3. To obtain the vibration information for detecting shaft cracks, mode identification XY proximity probes--located at various longitudinal positions along the rotor--and a Keyphasor reference are required. These probes make it possible to reliably observe the significant indicators of a shaft crack--the action of the shaft vibration patterns and bow changes at low rotative speed--as well as to identify nodal point regions and rotor mode shapes. The mode identification XY proximity probes should be used to continuously monitor machines that are susceptible to shaft cracking. The use of these probes also overcomes the potential danger, when only a single set of XY proximity probes is used, of locating the probes at a nodal point along the rotor. What to do when your machine experiences any of the shaft crack symptoms... As we stated before, because other malfunctions can cause the machine to exhibit similar symptoms as those experienced under a shaft crack condition, the proper methodology must be used to diagnose a shaft crack. Two recent shaft crack saves illustrate this point. The engineers originally suspected imbalance as the cause of the problem, but the machine did not respond properly to several balancing attempts. Difficulty in trim balancing often is a warning sign of a cracked shaft. Further analysis of the symptoms is required to determine the root cause of the problem. Bently Nevada possesses the capabilities to assist you in identifying the symptoms and detecting cracked shafts. We know numerous machines are now operating with at least small cracks and are using a three-pronged approach--research, services, and instrumentation--to detect shaft cracks. Research, conducted by Bently Rotor Dynamics Research Corporation, has resulted in the definition of shaft crack symptoms and the development of methodology for diagnosing shaft cracks. 4

THE FUNDAMENTAL ACTION OF A TRANSVERSE CRACK IS THAT IT I Orbit of shaft bow...... -- ~..... I...... I : I.' caused by transverse crack (1 X frequency BOWS THE SHAFT.... This causes: (1 Changing 1X (synchronous vibration behavior at speed and load. (2 Erratic response to attempted balancing. (3 Changes in the slow roll bow vector (at low rotative speeds. Further, as the bow increases, a totally different type of 2X behavior may occur. (See spectrum cascade of cracked shaft. Figure 1 5

2 3 4 5 6 7 8 RES kcpm TYPICAL ORBITS ~-\~ > c a: LU 0.. E ::1. 0. 0. ~-V I LU «u-t5j en 0.. ~ <t SPECTRUM CASCADE OF CRACKED SHAFT The above spectrum cascade plot shows the vibration response (amplitudes and frequencies from a proximity displacement transducer at different speeds during a shutdown of a rotor with a cracked shaft. Two types of 2X components caused by a cracked shaft can be observed in this example. When the rotative speed (800 rpm is at half the resonance speed (1600 rpm, the 2X frequency component has its resonance. The orbit shows the typical inside loop. This 2X motion is driven by a preload (like gravity and shaft nonsymmetry due to a crack. The 2X component is present even though the 1X component is small. When the shaft bow gets large as the rotative speed approaches the resonance speed a steady state preload causes the rotor reaction shown by the orbit at 1600 rpm. Notice that the 2X, 3X, and 4X harmonics are also incurred. This 2X motion is driven by the 1X and therefore, is only present when large 1X orbiting occurs. At 3600 rpm a large 1X vibration again occurs (second resonance. In conjunction with the preload, it causes the resulting 2X frequency component. The corresponding orbit is shown. Figure 2 6

TRAIN ONE POINT: BRG 1 VRT o TREND TYPE: 1 X AMP & PHASE -... "" '" 1 X... amplitude, \ and phase '\ vector \ \ \ 270 1----4--+--f---+--+-+--..l.--...L--4--------l 90 I I J / / /./' / /' /... --.._--- -/'./ 180 ROTATION The polar plot is a presentation of vector monitoring. The 1X (or 2XI amplitude and phase vector is monitored to ensure it remains within an acceptance region. When the vector moves outside the acceptance region, further diagnostics are necessary. This is especially important for monitoring for cracked shafts. Figure 3 Examples of orbits of a rotor with a cracked shaft and resulting shaft bow combined with a stead-state preload, such as gravity. Examples of orbits of a rotor with a cracked shaft where the 2X frequency component results primarily from a preload, like gravity. These are the classical 1X and 2X orbits for a cracked shaft, at a half first balance resonance speed. Figure 4 7

The methodolo is tau ht at our Customer Trainin seminars. rn-a mac ine save invo ving a s aft crac, at a pu ic service utility, engineers from the utility had attended a Bently Nevada seminar, which helped them better interpret the information from theil" 7200 Series Monitoring System and ADRE~ (Automated Diagnostics for Rotating Equipment system. Based on the vibration data, they stopped the machine. Inspection revealed a 400 degree spiral crack. The shaft crack methodo 10 lis Mechanica nglneerlng ervlces ln per customers. In a recent case, MES engineers' superior methodology made the difference in correctly diagnosing the problem. Bently Nevada's MES engineers analyzed the rotative speed (IX and the twice rotative speed (2X data and determined there was a high probability of a severe crack. The machine was disassenlbled. Inspection revealed a 90 to 95 percent crack through the rotor that was visible to the eye. Bently:Nevada's MES engineers are trained to detect shaft cracks and diagnose other machine malfunctions. They also are trained to perform balancing--using modern techniques--and to document the response vectors during the balancing process for the user's records. Several Bently Nevada instruments are available today for use in early shaft crack detection: o ADRE System to observe and document the machine response under transient conditions. The ADRE is especially useful for observing IX and 2X polar plots o 7200 Series Dynamic Data Manager system to continuously observe and trend IX and 2X amplitude and phase components as well as to observe and document spectrum information. o 7200 Series Monitoring System with the 72730 Digital Vector Filter CDVF-R to continuously observe IX amplitude and phase components. o Digital Vector Filter 2 CDVF 2 and Balance Master~ to take periodic readings of IX and 2X amplitude and phase components. o Keyphasor Multiplier, an accessory to the DVF 2 and Balance Master, which can be used to multiply the Keyphasor frequency to observe 2X components. o Snapshot to take periodic readings of IX and 2X amplitude and phase components and to plot data in spectrum, time base, and orbit formats. 8