CHAPTER 1 INTRODUCTION 1. 1 Introduction Condition Monitoring is a technique of monitoring the operating characteristics of plants, equipment or systems in such a way that changes in the monitored characteristics can be used to predict the need for maintenance before a serious deterioration or breakdown occurs. It aims at detecting condition leading to catastrophic breakdown and loss of service, reducing maintenance overhauls, fine tuning of operating equipment, increasing production and operating efficiency, minimizing replacement parts, inventory.
Condition Monitoring is very much essential for floating vessels like dredgers, where machinery like propulsion drive, dredge pump drive, jet pump drive etc. are subjected to high vibration levels during the process of dredging. The effective performance and the operational efficiency of a dredger depend upon the periodical maintenance of the dredging equipment and other components functioning in underwater and also the overboard fittings.
The maintenance of a dredger and its dredging equipment is highly expensive. By deferring or neglecting the maintenance function, the dredging operations may lead to be ineffective and the breakdown costs become uncontrollable. In the present work, Condition Monitoring using vibration analysis has been taken up on two trailing suction hopper dredgers. These two dredgers differ in the power drive units for propulsion and dredging machinery.
By implementing vibration monitoring to various key points of both the port and starboard sides on various systems like propeller shaft, propulsion gear box , engine shaft, shaft generator on propulsion drive; dredge pump shaft, dredge pump gear box, dredge pump motor shaft on dredge pump drive; jet pump shaft and jet pump motor shaft on jet pump drive; vibration parameters i. e. Root Mean Square (RMS) velocity and vibration echo levels are measured and probable faults are detected and suggestions to rectify the defects are made to improve the performance of the dredgers.
In the present work, the theoretical investigation has been carried out to monitor the vibration condition of propeller, dredge pump and jet pump shafts. The natural frequencies for the first five modes of these shafts are computed by Finite Element Analysis using Inverse Iteration Technique and Jacobi’s method. The frequencies are also verified using classical Rayleigh – Ritz method. For 1 an optimum rotational speed, the displacement and acceleration are calculated for each natural frequency. A comparison has been made between theoretical and measured values.
Frequency spectra for propeller shaft, dredge pump shaft and Jet pump shafts have been obtained using data collector. The vibration parameters have been measured by conducting experimental investigation onboard the dredgers using sophisticated instruments i. e. Data Collector, Integrating Vibration Meter and Bearing Echo Meter. The data has been obtained in the four quarters, each of duration three months in a year. Based on vibration parameters, various plots like overall trend plots for each point, average trend plots, exception ratio plots for each system and the whole drive have been obtained using a software, developed in C ++ anguage. From these plots, diagnostic analysis has been performed for suggesting probable causes of failure of each machine in each quarter and remedial measures have been suggested in the succeeding quarters of the year for improving the performance of dredgers. Finally, a computer system has been suggested for on line monitoring of the systems onboard the dredgers to increase the efficiency of the dredgers. 1. 2. 1. Introduction to Condition Monitoring Condition Monitoring has only recently come to the forefront of Maintenance Engineering as a further tool in the ongoing effort to maintain high levels of plant availability.
Condition Monitoring, in the opinion of Mcmahon, is the technique of monitoring the operating characteristics of plants, equipment or systems in such a way that the changes in the monitored characteristics can be used to predict the need for maintenance before serious deterioration or breakdown occurs. According to Stipho, maintenance in general aims at minimizing costs and maximizing the availability of the systems. The objective of Condition Monitoring as per Hensey and Nair is to maximize the performance of company’s assets by monitoring their condition ensuring that they are installed and maintained correctly.
Collacott states that the act of Condition Monitoring has actually been practiced by plant engineers for generations, estimating how long an item of equipment can continue in service until it needs to be shut down for repair, using the traditional touch, see and hear methods. However the weakness of traditional systems is its reliance on the 2 experience and integrity of individual plant engineers, undertaking the assessment and his familiarity with the potential problems of particular item of the equipment.
Condition Monitoring is a novel technique of detecting in advance any incipient failure with ease and confidence in any part of the dynamic systems. Unless there is a holistic practice to monitor, diagnose and prognose the undesirable symptoms in every part of the systems as mentioned by Wowk, it will be difficult to keep up with the rapid pace of development technology. Jones says that Condition Monitoring can also be a test and quality assurance system for continuous processes as well as discrete component manufacture.
It detects condition leading to catastrophic breakdowns and loss of service, reducing maintenance overhauls, fine tuning of operating equipment, increasing production and operating efficiency and minimizing replacement parts inventory. The application of computers, electronic measuring and detecting systems has provided a new impetus to Condition Monitoring. The maintenance methods can be broadly classified into the following three methods i. e. breakdown, preventive and predictive maintenance methods. In the past, machines were often kept running until they fail. This caused costly downtime to fix or replace the defective machines.
Often, failure of one machine damages other machines resulting in even more downtime and higher repair costs. This method is known as “Breakdown Maintenance” or “Run-to-failure Maintenance”. Till recently, machinery failure was avoided using “Preventive Maintenance”. This involves scheduling periodic downtime for visual inspection of parts or to replace equipment regardless of its condition. This method will be very expensive since equipment is removed or repaired in spite of having useful life. “Predictive Maintenance” has been developed to get the maximum life out of machines without allowing the machines to fail.
It involves monitoring machines to predict when a machine is likely to fail, thus allowing to schedule repairs or replacements. This maximizes the life of machinery without allowing any breakdown. Neale and Woodley emphasized the two main methods i. e. Trend Monitoring and Condition Checking used for Condition Monitoring. Trend Monitoring is the continuous or regular measurement and interpretation of data collected during machine operation to indicate variations in the conditions of the machine or its components, in the interest of safe and economical operation.
This involves the selection of some suitable and measurable indication of machine or component deterioration, such as one of those listed in Fig. 1. 1 and the study of 3 the trend in this measurement with running time to indicate when deterioration is exceeding a critical rate. The principle involved is illustrated in Fig. 1. 2, which shows the way in which such trend monitoring can give a lead time before the deterioration reaches a level at which the machine would have to be shut down. This lead time is one of the main advantages of using trend monitoring rather than simple alarms or automatic shutdown devices.
Figure 1. 1: The Indications of Machine or Component Deterioration (Ref. 13) 4 Figure 1. 2: The Regular Monitoring of Deterioration to Give Advanced Warning of Failure (Ref. 13) Condition checking is where a check measurement is taken with the machine running, as a measure of the machine condition at that time. To be effective, the measurement must be accurate and quantifiable, such that these values will not exceed the limiting values that are set for standardizing the running hours. These two methods of condition monitoring are compared in greater detail in Table 1. , and the resulting advantages in terms of the provision of lead time and better machine knowledge are shown in Table 1. 2. 5 Table 1. 1: A Comparison of Methods of Condition Monitoring and Failure Diagnosis (Ref. 13) 6 Table 1. 2: Advantages of the Use of Condition Monitoring (Ref. 13) There are a number of valid and effective methods available to the engineer as mentioned by Allenby. It is vital that careful consideration to the method be given to obtain the most effective pay back in shortest possible time.
Another important factor is that the method can be readily interpreted and understood by all people involved in this commitment. The choice of the parameters to be monitored and the measurement techniques has a tremendous impact on the efficiency of the condition monitoring philosophy. Vibration monitoring, is a well-established method for determining the physical movement of a machine or structure, due to imbalance mounting or alignment. This method can be obtained as simple, easy to use and understand, or as a sophisticated real time analysis.
All rotating and reciprocating machines vibrate either to a smaller or to a greater extent. Machines vibrate because of defects or inaccuracies in the system. When the inaccuracies are more, it results in increased vibration. Each kind of defect produces a vibration characterized 7 in the unique way. Therefore recording vibration level of a machine indicates the condition of the machine. Debris analysis is well proven in all types of industry and works on the principle of taking a known quantity sample from, for example, a gear box, then analysing the amount and type of foreign particles present in the sample.
This will show such problems, as gear wear, if the sample detects particles of gear material. Oil analysis differs from Debris analysis in so far as this technique allows an assessment of the actual condition of the oil in use that is, whether the oil quality good enough for the application after a period of use, or it is burnt or exceeded its useful life. Corrosion monitoring, is usually applied to the fixed plant containing aggressive materials to monitor the rates of internal corrosion of the walls of the equipment. It is the systematic easurement of corrosion or degradation of an item of equipment, with the aim of assisting in understanding of corrosion process or obtaining information for the use of controlling corrosion or its consequences. Thermography is a rapidly developing technique. It provides, via colour camera’s and video’s, clean indications of heat losses, hot spots, cold spots such as switch gear or any piece of plant or production processes, where temperature or its effect is important. It can be used both as maintenance tool or a quality assurance tool.
Shock pulse method is a unique technique for monitoring the true operational condition of a bearing by measuring the pressure wave generated by an instantaneous mechanical impact. Visual monitoring, involves inspection and recording of surfaces to detect defects such as surface cracks and their orientation, oxide films, weld defects and the presence of potential sources such as sharp notches or misalignment. Analysis is the most important phase of condition monitoring to establish the maintenance schedules with confidence.
The computer systems keep database of fault frequency parameters, designs of equipment etc. and thus allow easy fault frequency identification. Computers can be used to subtract spectrum from each other, perform spectral analysis and perform other calculations, display the data in convenient formats. Trend changes in the fault frequencies can be reviewed for forecasting the future levels, comparing with the alarming levels. The main benefit of computer is that it is able to quickly compute great number of parameters based on the original data, which can highlight different aspects of the data and 8 an compare/check with the alarming levels, set for each machine. Computers facilitate the following three types of reports. Simple report is listing the structure of data base, enables the operator to quickly determine what measurements have been collected. Professional Format produces a graph describing the results of the analysis and maintenance recommendations along with a diagram of machine in question. Maintenance Management Report informs the operator about the condition of the plant, highlighting the machines that require attention.
Computer based systems provide fast and accurate generation of reports, which form vital link between the standard condition monitoring functions and plant maintenance. 1. 2. 2. Vibration Based Condition Monitoring Vibration based condition monitoring was developed over 30 years ago to detect and diagnose faults in critical rotating machinery, the objective being to reduce the probability of breakdown. Peter says that the vibration based condition monitoring programme is governed by three basic facts: a) It is normal for machines to vibrate; the perfect machine has yet to be made. ) The onset of machinery problems is usually accompanied by an increase in vibration. c) Each rotating element generates its own unique vibration frequency. An engineer, trained in vibration analysis techniques, can determine the operating condition of a machine in a similar manner as the doctor assesses the heartbeat of a patient with a stethoscope. By adopting such a sound principle, machinery faults such as unbalance, misalignment, looseness, worn gears, defective bearings, electrical faults, aerodynamic forces etc. , can be readily diagnosed while a machine is still operating.
Due to the world-wide acceptance of this method as an essential maintenance tool and advances in high technology, the maintenance manager is now faced with a wide choice of monitoring systems. These include the simple check instrument with clip board and pencil for manually recording measurements, microprocessor controlled data acquisition units, computerized data management systems and machinery information centers. In order to achieve the full benefit of a condition monitoring programme a company’s maintenance strategy and operating procedures must first be established.
Then, depending on the size and/or location of the process plant, the maintenance manager should select a system 9 to suit the immediate requirement which must also be flexible enough to accommodate the longer term needs. The success of the chosen system will depend on its ability to generate useful information rather than masses of raw data. Further, to ensure that it integrates with the process plant, both maintenance and production personnel should be involved from initiation through to routine operation. Condition Monitoring has found great favor in the transport, energy and process industries, e. . , aerospace, marine, power generation and oil/gas production. The machines used in these areas are generally well trained and tested and are fairly categorized. The potential benefits of this technique are enormous, particularly for continuous processes where high cost equipment and ancillaries are required to operate at both maximum efficiency and near the design limit. After changing from calendar based maintenance to condition monitoring based maintenance, as per the case studies on floating vessels conducted by Hind, M.
D & Robinson W. D reduced maintenance man hours by 37%. Table 1. 3: Man Hours Reduced by Condition Monitoring Based Maintenance as Compared to Calendar (Ref. 16) 10 1. 2. 3. Condition Monitoring Of Rotating Machinery Vibration and noise are invariably produced when an element of a machine is rotating. In recent years, there has been considerable interest in the maintenance techniques with the help of condition monitoring of rotating machinery based on the analysis of vibration characteristics generated by machinery.
The technique is used in conjunction with on-line maintenance that is maintenance of machinery when conditions indicate that a problem is about to occur and warrant a correction. Although the condition monitoring and analysis will increase the operational safety level, it will not be able to prevent failures. What it will do is to enable the progressive deterioration of a machine to be detected and for appropriate remedial action to be initiated.
Existing standards of establishing vibration limits for rotating machinery were derived from experience and subjective opinions with little regard to the dynamics of a particular machine as quoted in VDI 2056 . Firoozian observes that the short comings of such criteria have been pointed out since 1970’s. The dynamic behavior of shafts rotating in journal bearings as stated by Eshleman has attracted an enormous amount of research interest in the past three decades and have addressed such problems as critical speeds, rotor stability and its responses, non-linear behavior of the lubricating oil film etc. But from condition monitoring point of view Morton observes that the literature is less extending and primarily concerned with detecting changes in operating conditions. There is a growing awareness within production and maintenance circles of the benefits to be enjoyed from vibration based condition monitoring applied to rotating shafts as observed by Allwood to increase reliability and accuracy of the monitoring techniques. The parameter identification methods offer potential benefits as suggested by Stanway . These methods have been successfully used in active and adaptive controls of complex machinery.
In this thesis, the shaft drives under investigation produce the cyclic motion, which is purely rotational; the main cause of vibration is due to the geometric axis of rotation seldom coinciding with principal axis of inertia of the rotating part. The lack of this concentricity causes vibration. In addition to this, misalignment of shafts, wear of parts, imbalance in rotors, and increase in clearances may also contribute significantly to the vibrational behavior. These vibrations are objectionable, which produces heavier loading on parts such as bearings and 11 epeated stresses all along the line which ultimately may result in the fatigue failure and so every effort must be made to remove or greatly reduce them. A computational assessment of the vibration behavior of the machines is of significance with regard to the optimization of the complete installation. The primary use of condition monitoring is derived when the end user knows whether to tighten a bolt, rebalance the main shaft, replace a bearing etc. In order to identify the fault, it is necessary to examine the vibration spectrum. Each physical part of the machinery will vibrate at different frequency.
The main shaft rotates at fundamental frequency, if there are six blades, the vibration will occur at six times the fundamental frequency. Rao observed that due to the demands of high speed operation and the use of light structures in the turbo machinery, dynamic measurements are necessary and vibration testing has therefore found wide spread use. The basis for such a health monitoring is to make some relevant calculations together with practical measurements. Ruhl and Robert et al. [25,26] states that Finite Element Method is more effective in computing natural frequencies of rotating shafts.
Timoshenko observes that geometry boundary conditions are more effective in such method. Ramamurti presented a few solution techniques to compute the smallest or largest number of natural frequencies and natural modes of the shafts. Even though, main shafts in the dredging machinery are made of high strength steel, they deflect a small amount. This distortion creates forces on nearby bearings and sets both the structures into cyclic motion. According to Brown the bearings create reactionary forces which prevent the shaft from moving as much it would tend to.
These reactions prevent the normal sine wave motion of the shaft from achieving its full excursion in amplitude. In other words the sine wave motion of the rotor deflection is distorted at the extremes resulting in harmonics, according to Zhu and Firoozian  the simplest form is the regular recording of relevant parameters in a “Log”. Regular logging of important parameters and their derivatives compared with the reference information provides the basis for performance trend monitoring. Velocity is the best indicator of vibration intensity   . According to Hewlett-Packard , ibration monitoring includes both narrow band frequency and trend analysis of overall vibration levels. Narrow band analysis is best suited for analysing signals containing discrete tones, example machine vibration, where as a constant band width gives uniform resolution on a linear frequency scale and gives equal resolutions and separation of harmonically related 12 components which facilitate detection of a harmonic pattern. The constant percentage band width as observed by Burrows gives uniform resolution on a logarithmic frequency scale and this can be used over a wide frequency range.
Trending overall vibration levels as given by Neale  is easy to measure and useful for identifying several vibration problems. The trending programme as shown in Fig. 1. 3 will take the form of a pyramid. At the top of the pyramid are a few machines with serious defects that require close scrutiny. Next are those that require detailed diagnosis because of an upward trend or other irregularity. Below that are the machines that are routinely trended. At the bottom are the large number of good machines that may or may not have vibration data on record but are not monitored on a regular basis.
This segregation is desirable to focus attention on those machines that need attention. The biggest saving to be gained from a vibration Monitoring programme is avoiding losses due to unexpected breakdown. Figure 1. 3: Arrangement of Machines in Trending Programme (Ref. 34) 13 CHAPTER2 THEORETICAL ANALYSIS 2. 1 INTRODUCTION The Dredger- A and Dredger – B under investigation for Condition Monitoring and Diagnostic Analysis are described in detail. Theoretical investigation of shaft drives of propulsion and dredging machinery using Finite Element Method has been presented.
Natural frequencies of the torsionally equivalent shafts are computed using Inverse Iteration technique and Jacobi’s method. The frequencies are compared with those obtained by Rayleigh Ritz method. Acceptable values of displacement and acceleration as per vibration standards of vibration velocity have been computed for each frequency such that the shaft drives are stable. Mode shapes at certain intermediate frequencies are also furnished for assessing the vibration behavior of the shafts. 2. 2. 1 PROPULSION SHAFT DRIVE
The drive consists of controllable pitch propeller, propeller shaft, gear box, main engine and other auxiliary components like seals, bearings etc. , on both the sides of the vessels i. e. , port side and starboard side. The propeller shaft is surrounded by stern tube to arrest flow of water into the engine room. The shaft is a hollow bored one and coupled to the pitch actuating mechanism, by means of a loose shaft coupling flange. The photographic view of whole drive on port and starboard sides for Dredger -A and Dredger – B, indicating various systems are shown in Fig. 2. 1 and Fig. 2. 2 respectively. 14 15 Figure 2. : Propeller Shaft Drive of Dredger A 16 Figure 2. 2: Propeller Shaft Drive of Dredger-B 2. 2. 2 DREDGE PUMP DRIVE This drive contains dredge pump, pump shaft, gear box, motor and other auxiliary components like couplings, bearings etc. The Dredger-A is equipped with dredge pump and jet pump being driven by separate motors where as in the Dredger -B they are driven directly by the main engine and hence it does not have a separate motor. This drive is also provided on both the sides i. e. Port side and Starboard side of the vessels. Fig. 2. 3 and Fig. 2. 4 show the photographic views of drives of Dredger -A and Dredger- B in detail.
Figure 2. 3: Dredge pump shaft drive of Dredger-A 17 Figure 2. 4: Dredge pump shaft drive of Dredger-B 2. 2. 3 JET PUMP DRIVE Jet pump drive contains jet pump, motor and other auxiliary components like couplings, bearings, etc. This drive also does not have a separate motor. In Dredger – B as jet pump is directly driven by main engine. The photographic views of drives of the Dredger – A and Dredger- B are indicated in Fig. 2. 5 and Fig. 2. 6. 18 19 Figure 2. 5: Jet Pump Shaft Drive of Dredger-A 20 Figure 2. 6: Jet Pump Shaft Drive of Dredger-B 2. 3. COMPUTATION OF MODAL PARAMETERS
The natural frequencies, computed using the Raleigh-Ritz method and FEM are given in Table 2. 2. Mode shapes obtained for shafts at some intermediate frequencies are shown in Fig. 2. 7 to Fig 2. 16. These natural modes illustrate the shape of each shaft in the selected frequency range. Figure 2. 7: Mode 1 Shape of Propeller And Dredge Pump Shafts Figure 2. 8: Mode 2 Shape of Propeller And Dredge Pump Shafts 21 Figure 2. 9: Mode 3 Shape of Propeller And Dredge Pump Shafts Figure 2. 10: Mode 4 Shape of Propeller And Dredge Pump Shafts 22 Figure 2. 11: Mode 5 Shape of Propeller And Dredge Pump Shafts Figure 2. 2: Mode 1 Shape of Jet Pump Shaft 23 Figure 2. 13: Mode 2 Shape of Jet Pump Shaft Figure 2. 14: Mode 3 Shape of Jet Pump Shaft 24 Figure 2. 15: Mode 4 Shape of Jet Pump Shaft Figure 2. 16: Mode 5 Shape Of Jet Pump Shaft In order to evaluate precisely the vibration pattern at all frequencies, it is recommended to refer the vibration velocity which is decisive of vibration intensity. When displacement is considered, vibrations at lower frequencies are overvalued. In contrast, if acceleration is considered, this would over value the very high frequencies of vibration but not detect the low frequency 25 ibration. In order to ascertain the vibration behaviour of shafts at low as well as high frequencies, acceptable values of displacement and acceleration are computed at each mode for the trouble free velocity (0. 5 ;12. 7 AA Extremely rough, dangerous; Shut down Very rough; correct within few weeks; check monitor frequently Rough; correct to save wear as soon as possible Fair; minor fault; uneconomical to correct Smooth, well balanced; well aligned 0. 3 – 0. 5 7. 6 – 12. 7 A 0. 2 – 0. 3 5. 1 – 7. 6 B 0. 1 – 0. 2 2. 5 – 5. 1 C