On-Site Application of Equipment Condition Monitoring and Fault Diagnosis Technology
—— A Brief Analysis of the Relationship Between Traditional Diagnosis, Simple Diagnosis, and Precise Diagnosis
Author: Xiao Ze
Chapter One: Introduction
Section One: Forum Impressions
I have been logged into the vibration forum for several months now, and overall, I feel that this is a harmonious, innovative, and convenient learning and communication space. Through learning and communication, I have greatly benefited, broadened my horizons, and improved my level. The vibration forum is a great place for modern fault diagnosis personnel, especially for on-site diagnostic personnel, to engage in self-learning and continuing education.
Through visiting the forum, I also feel that there are still some shortcomings in the thinking and understanding of newcomers to accident fault diagnosis. Therefore, in this lecture, I do not want to reiterate too much of the technical theories and methods that everyone is already familiar with, but rather I want to start from the practical aspects of diagnostic work on-site and discuss some conceptual ideas with everyone, hoping to form a complete on-site work thought and system in everyone’s mind. This has important and practical significance for the overall development of on-site fault diagnosis work and improving the overall quality of professionals.
Finally, I will provide some classic case analyses. Through these case analyses, everyone can feel the ideas proposed in the lecture and should not simply understand these cases as technical problems or successful experiences in diagnosis.
Having engaged in fault diagnosis work for more than 20 years and read a large amount of data and textbooks, I have generated many thoughts, especially from the on-site practical perspective, leading to new understandings of certain past concepts or definitions, which I will present for discussion. Some of these understandings may not necessarily be correct or accurate.
New Concepts of Condition Monitoring and Fault Diagnosis
1. Equipment Condition Monitoring
Condition monitoring technology based on monitoring the vibration development trend of equipment.
It can be seen that condition monitoring is based on monitoring, aimed at determining the operational status of equipment, and targeting forecasting the vibration development trend of equipment. By long-term monitoring of the most basic vibration data (such as displacement, velocity, acceleration), and according to international or national simple diagnostic vibration standards, it determines whether the equipment vibration exceeds the standard or forecasts the operational trend of the equipment. Mathematical methods are used to estimate when the operating equipment may exceed the standard vibration level, necessitating a shutdown for maintenance; the specific analysis of the fault causes, locations, and nature is not the main task. Of course, it is still somewhat difficult to accurately determine the fault’s location, cause, and nature solely based on condition monitoring.
2. Equipment Fault Diagnosis
Fault diagnosis technology based on analyzing the vibration characteristic parameters of equipment.
It can be seen that fault diagnosis is based on analyzing the vibration characteristics of equipment, aimed at determining the fault location, cause, and nature, and targeting ensuring the economical operation and scientific maintenance of the equipment.
Fault diagnosis mainly involves a more detailed analysis of abnormal (faulty) equipment, aiming to identify the characteristic parameters corresponding to the fault, achieving the goal of identifying the fault location, cause, and nature. It is a precise diagnosis built upon simple diagnosis, rather than a standalone process that disregards simple diagnosis, which we will discuss in detail later.
Determine the operational status and development trend of the equipment;
Analyze the causes, locations, and nature of equipment faults;
Provide scientific maintenance basis, indicating the direction for repairs;
Reduce maintenance and upkeep costs, enhancing the economic efficiency of equipment operation;
Ensure production safety, providing decision-making information;
Section Four: The Significance of Conducting Equipment Fault Diagnosis Work
Conducive to improving equipment management levels;
Avoid major equipment accidents and reduce the severity of accidents;
Potentially gain significant economic and social benefits
Section Five: Classification of Monitoring Methods
1. Continuous Monitoring: Also known as online monitoring, it is a precise diagnosis using data collection and computer analysis technology, including remote fault diagnosis technology. Advantages: Comprehensive information collection, rich analysis methods, and high accuracy. Disadvantages: High equipment investment, and operators need a high theoretical foundation.
2. Periodic Monitoring: Conducted at determined time intervals, generally using simple, portable detection instruments, belonging to simple diagnosis. Advantages: Simple equipment, low investment, easy to operate. Disadvantages: Information collection and analysis are relatively simple.
3. Fault Monitoring: Based on inspections by operators and maintenance personnel, testing and analyzing the equipment when abnormal operation is detected, seeking fault causes, and assessing operational conditions. Due to advancements in detection instrument hardware technology, traditional periodic condition monitoring methods have been largely abandoned, making this the most commonly used monitoring method on-site, suitable for small units or offline monitoring equipment, but not applicable for large units, especially those with sliding bearings.
Section Six: Classification of Faults
There are many classifications of equipment faults in various materials and textbooks, with detailed classifications from various perspectives. For example, local faults vs. overall faults, sudden faults vs. gradual faults, etc., but there is a lack of classification from the perspective of vibration analysis. Based on 20 years of on-site work experience, I propose for the first time to categorize equipment faults into: Endogenous Faults and Exogenous Faults.
1. Endogenous Faults: Faults caused by defects in the equipment’s own components leading to abnormal vibrations. These faults are unrelated to external conditions of the machine; they are caused solely by internal reasons. For example: defects in parts; installation defects; improper fits between components, etc.
2. Exogenous Faults: Faults caused by changes in external conditions or parameters leading to abnormal vibrations. These faults are not directly related to the machine’s manufacturing, installation, or design, but are simply caused by changes in external conditions. For example: temperature, pressure, vacuum, overload, improper operation, etc.
Chapter Two: Traditional Fault Diagnosis
Section One: The Meaning of Traditional Fault Diagnosis
所谓“传统故障诊断”就是我们常说的感官诊断。主要是依靠人的“眼看、耳听、手模、鼻子闻”，通过人体的感觉器官来感知设备的振动、温度、声音、气味等，获得这些物理特征的模糊量值，达到主观判断设备运行状况的目的。所以传统诊断是以模糊量值为样本的，不能够定量地反映设备的各种物理特征，具有模糊定性的性质。
Section Two: Methods and Significance of Traditional Fault Diagnosis
Traditional diagnosis is the most practical and direct application of fault diagnosis technology from the field. It is called a technology, not just an experience, because it has its own characteristics and inherent aspects that need to be studied and mastered, which is not something anyone can casually pick up.
Traditional diagnosis’s technical content is, of course, more based on experience, but it is not merely based on experience; it has incorporated modern diagnostic ideas and methods, forming a unique diagnostic technology on-site, while purely experiential diagnosis should be called “craftsman diagnosis”.
In traditional diagnostic methods, the most involved is listening: how to listen, and what part to listen to?; secondly, there is feeling: perceiving the strength of vibrations and the temperature through touch, also involving how to touch and which part to touch; and there is also seeing: how to look and where to look is also a question. In the field, it is often encountered that: how come I didn’t hear it? How come I didn’t see it? So I ask: why did others hear it? Why did they see it? Therefore, I say traditional diagnosis still has its technical aspects.
For example: listening, everyone can hear, but knowing what part to listen to requires a deeper understanding of the equipment’s structure, performance, and working methods; achieving diagnostic purposes through listening is not something everyone can do. By distinguishing sounds, one can generally judge the condition of bearings, gears, and even discern whether the planetary gear’s pin is broken.
When judging the condition of bearings, we mostly listen to the sounds from the bearing area, distinguishing the pitch of the sound, determining whether there are any additional abnormal sounds superimposed upon the background noise generated by the bearing’s rotation; when judging whether the rotor has a scuffing issue, we listen more to the area between the rotor and the stator casing; when discovering a similar knocking sound in the background noise of the motor, we listen to the motor casing; when judging whether the pump has a cavitation issue, we often listen to the pump’s inlet piping…
When judging pump vibrations, we can feel the magnitude of vibrations in both the horizontal and vertical directions through touch, aiming to discern whether there is a foundation looseness issue; when assessing system rigidity, we feel the vibrations of the foundation frame in both directions; when judging temperature, we use the fingertip of our hand to probe, and then formally sense the temperature…
Eye observation diagnosis: I will give a practical example: In September 2006, workers in the water supply workshop discovered during routine checks that a large water pump’s motor was vibrating significantly, notifying equipment management personnel for vibration testing. The testers found that the instrument measured very low vibration (almost no vibration) with no signs of fault, but on-site observation and touch indicated significant vibration, leading to doubts about the instrument’s functionality, and replacing the instrument yielded the same results.
Upon arriving at the site and carefully observing the equipment’s vibration condition, I concluded: Due to insufficient rigidity of the support system, the motor vibrated significantly. Everyone was puzzled, and I explained my analysis as follows:
Through auscultation of the motor, there were no abnormal sounds from the bearings or rotor; the process parameters were normal, and the instrument measured almost no vibration (displacement, velocity, acceleration), but the visual and tactile observations indicated clear vibration, especially as the visual observation could discern that the motor casing was oscillating horizontally, indicating a very low vibration frequency, while the instrument used an accelerometer, which has a lower frequency limit of 10Hz. Therefore, when the vibration frequency is below the sensor’s frequency limit, it cannot properly pick up the signal, resulting in a very low vibration measurement. Additionally, upon observing the motor’s support foundation, it was noted to be a thin steel frame structure with insufficient rigidity. This is a typical case analysis from traditional diagnosis.
Chapter Three: Simple Fault Diagnosis
Section One: The Meaning of Simple Fault Diagnosis
Simple diagnosis primarily relies on small, simple instruments to measure the vibration conditions of equipment, measuring the simplest physical characteristic quantities, mainly the effective values of vibration velocity, displacement, and acceleration, and determining whether the equipment requires maintenance based on comparing the effective vibration velocity against international or national vibration standards.
Section Two: Methods and Significance of Simple Fault Diagnosis
Simple diagnosis is based on measuring simple vibration parameters of equipment, but although the measurement data is simple, it has significant meaning in on-site applications. The simple data reflects a wealth of fault information. Below, I will introduce them separately:
1. Vibration Displacement: Most instruments measure peak displacement. From displacement, combined with velocity and acceleration measurements, we can determine whether the vibration originates from low frequency or high frequency; from the magnitude of displacement in both horizontal and vertical directions, we can preliminarily judge whether there is a foundation looseness issue; from displacement, we can also analyze the rigidity issues of the equipment foundation…
2. Vibration Velocity: Generally, simple diagnostic instruments measure the effective values of vibration velocity. This is consistent with international or national vibration standards, and using the vibration velocity measurement compared to relevant standards can determine the basic operational status of the equipment (good, usable, unusable), which has significant guiding importance for extending equipment operating time and improving operational efficiency.
The measurement of vibration velocity also serves a very important function in the acceptance of the quality of rotating equipment purchased or repaired. Because international and national standards, including some industry standards, have clear vibration requirements for rotating equipment, a series of regulatory standards have been established that any enterprise must follow. For example: ISO-2372, GB10068-2000, etc. Users and suppliers need to establish a common accepted inspection platform to determine whether the equipment quality meets standards, and vibration testing is a simple and feasible acceptance method.
For example: When you receive an electric motor, general enterprise users do not have specialized motor testing equipment. How to judge whether the motor quality is acceptable can be based on the national GB10068-2000 motor vibration quality standard for acceptance. For general motors, when the center height is within 400mm, freely suspended, and the speed is between 600~3600rpm, the vibration must not exceed 2.8mm/s (the international standard is 3.5mm/s). This establishes a unified acceptance standard platform, effectively preventing manufacturers from deflecting responsibility when vibration issues arise.
3. Vibration Acceleration: Measuring acceleration can determine whether the equipment has high-frequency vibrations, especially for analyzing rolling bearing faults, it has very practical significance on-site.
4. Other Useful Information in Simple Diagnosis: During the process of simple diagnosis, it is recommended to use instruments that indicate simulated output. This is because digital display instruments, during A/D conversion, lose some information, whereas the display of analog meters is real-time, unaffected by sampling periods, thus having an inherent advantage in observing vibration fluctuations (impacts). For instance: if the bearing race is not circular, causing rolling elements to momentarily jam and experience slight friction, the situation of the inner or outer race of the bearing “playing loose” can be easily distinguished.
5. Simple Diagnostic Instruments Already Have Basic Frequency Analysis Functions: This is very practical on-site. For general faults, the frequency analysis function can fully rival that of precise diagnostic instruments.
Chapter Four: Precise Fault Diagnosis
Section One: The Meaning of Precise Fault Diagnosis
Precise fault diagnosis mainly relies on computer technology, mathematical processing technology (including FFT, IFFT), data acquisition technology, frequency analysis, and other techniques to analyze the vibration conditions of equipment, measuring and collecting the physical quantities of vibration, and through mathematical processing, extracting vibration characteristic parameters to seek and determine the causes, locations, and nature of vibration faults, providing reference for fault elimination.
Section Two: Methods and Significance of Precise Fault Diagnosis
Precise diagnosis is the highest practical result of the current development of vibration fault diagnosis. It provides a very effective tool for fault diagnosis personnel, helping them improve the accuracy of fault judgment. Especially in the online analysis monitoring of large units, the technical advantages are highlighted, playing an irreplaceable role in early detection of fault signs, preventing catastrophic accidents in large units, and ensuring safe production.
The most widely used and mature technology is frequency analysis, which everyone has a strong sense of. In fact, in on-site applications, the position of the shaft and the trajectory of the shaft are also very meaningful, particularly for the dynamic pressure sliding bearings of large units, as the change in the shaft position and the shape of the shaft trajectory directly reflect the stability of the rotor, effectively preventing oil film whirl and avoiding oil film oscillation. Compared to frequency analysis, it allows professionals to pay sufficient attention before the equipment generates whirl, thereby avoiding its occurrence.
Currently, most online monitoring systems have remote diagnosis capabilities, facilitating remote diagnosis or expert consultations. The development of computer broadband networks has laid the foundation for remote diagnosis.
Chapter Five: The Relationship Between Traditional Diagnosis, Simple Diagnosis, and Precise Diagnosis
Section One: The Concept of “Trinity”
From the perspective of practical work on-site, the relationship between traditional diagnosis, simple diagnosis, and precise diagnosis is complementary; there is no issue of which is advanced or backward, nor can any replace the other. They constitute a “trinity” on-site fault diagnosis work system. Each has its own advantages and disadvantages, strengths and weaknesses. Based on the characteristics and relationships between the three, establishing a complementary diagnostic work model is feasible and necessary. Therefore, I propose the idea of “trinity” for the first time.
The auscultation and observation in traditional diagnosis cannot be replaced by other diagnoses; simple diagnosis has become simple and feasible in overall comprehensive judgment of equipment operational status due to international and national standards, especially having an absolute position in equipment quality acceptance; precise diagnosis holds a crucial position in analyzing and determining the causes, locations, and nature of faults, especially in the operational monitoring of large units.
Section Two: The Working Significance of “Trinity”
On-site diagnosis should be simple when possible; for general faults, small instruments can solve them. Many of my later cases are analyzed and diagnosed with small instruments. Of course, for some large equipment or complex issues, precise diagnosis must be performed.
Using small instruments often better enhances human quality; from the perspective of training and cultivating personnel, small instruments have advantages. Please do not misunderstand; my deeper meaning is that many people engaged in diagnostic work, due to the development of instruments and technology, have fewer opportunities to engage with and use simple and traditional diagnoses, which is a deficiency in cultivating diagnostic talent.
An outstanding on-site fault diagnosis engineer must experience learning and practice in traditional diagnosis, simple diagnosis, and precise diagnosis; lacking any one of these links would be incomplete. From my personal experience, traditional diagnosis and simple diagnosis are more challenging, but the knowledge gained from them is far richer and more solid than that from precise diagnosis. Without the experience of traditional and simple diagnosis, you may not be able to communicate with maintenance personnel on-site, creating a language and professional distance that may prevent you from understanding more real situations. After establishing the most fundamental knowledge, applying precise diagnosis will be a natural progression.
For those who start with precise diagnosis, I suggest you quickly fill in the missing courses; otherwise, you can only become an ivory tower “scholar”.
In the actual on-site diagnostic work, the first step is traditional diagnosis to obtain necessary non-quantitative vibration information. Based on this information, one can form a preliminary sensory impression of the fault’s cause, location, and nature, and evaluate the operational status of the equipment. Then, simple diagnosis (including basic frequency analysis) is conducted to determine whether the vibration exceeds the standard, obtaining simple vibration parameters and the main vibration frequencies (note: current simple diagnosis is no longer the traditional simple diagnosis, but is between simple and precise), referencing the preliminary impression from traditional diagnosis. If the two align closely, one can basically determine the fault’s cause, location, and nature, and draw conclusions.
Conversely, if the information reflected by the instruments does not align with the preliminary impression from traditional diagnosis, or if there are significant discrepancies, one must return to traditional diagnosis, re-sensing based on simple diagnosis to see if the information obtained can be reasonably explained; if it can be explained, a diagnostic conclusion can be drawn.
If unification cannot still be achieved, precise diagnosis begins, employing more advanced and fully functional diagnostic equipment to obtain more detailed vibration characteristic information, forming diagnostic impressions (not conclusions). Based on this diagnostic impression, one attempts to explain the information obtained from traditional and simple diagnosis, integrating information from all three stages to explain, verify, complement, or even correct, aiming for a high degree of unity.
Once unification is achieved, diagnostic impressions can be transformed into diagnostic conclusions; otherwise, if a definitive or complete conclusion cannot be provided, the uncertain and incomplete parts may require actual disassembly of the equipment or further monitoring to supplement necessary information to form new, more certain and complete conclusions.
At times, after practical verification or repair validation, diagnostic conclusions may be incorrect, necessitating careful summarization of lessons learned to identify the cause of the errors. As long as the cause is clear, experience is gained, and lessons are accepted, you progress and improve; from this perspective, you have not failed!
In on-site diagnostic work, it is essential to organically combine traditional, simple, and precise diagnoses based on actual conditions, consciously applying the “trinity” concept to maximize effective information acquisition, comprehensively summarizing information from all three levels to accurately judge equipment faults, ensuring safe production, and providing scientific maintenance basis and direction.
Chapter Six: Key Issues to Note in Fault Diagnosis
Section One: Major Taboo in Equipment Fault Diagnosis
Making conclusions based on a single spectrogram
Focusing on the spectrogram while neglecting the on-site conditions and machinery
Overemphasizing the correlation between characteristic frequencies and faults
Following textbook “rules” without proper verification
Failing to discern the authenticity of signals
Neglecting the patterns of state changes
Ignoring information related to vibration
Only focusing on modern methods while neglecting traditional experience
Starting solely from experience and rejecting modernization
Viewing the mountain from different angles yields different conclusions;
The heights and distances vary across perspectives.
Not recognizing the true face of Mount Lu
Only because one is in this mountain.
Different observational angles yield certain differences in conclusions; thus, emphasizing information integration is key to correctly solving problems.
Below, I will provide a simple personal summary of the information that on-site diagnostic personnel should pay attention to and grasp, not targeting a specific piece of equipment but analyzing as many devices as possible, suggesting everyone note these when going on-site.
1. Basic Equipment Information
① Equipment model and nameplate parameters: such as motor grade, voltage, current; compressor speed, critical speed, etc.
② Basic structure, performance, and purpose of the equipment: such as whether the foundation is concrete or steel frame; whether the rotor is cantilevered, single-stage, or multi-stage; the number of impeller blades; whether it has variable frequency speed regulation; working medium, sealing form, etc.
③ Process parameters: such as process medium, flow, pressure, temperature; type of lubricating oil, oil pressure, temperature, etc.
2. Equipment Bearing Types
① Rolling bearing types: deep groove ball bearings, angular contact bearings, cylindrical roller bearings, tapered roller bearings, pure axial thrust bearings; rolling elements can be single-row or double-row. It is best to have the bearing model.
② Sliding bearing types: static pressure sliding bearings, dynamic pressure sliding bearings, or servo-controlled bearings; among dynamic pressure sliding bearings: oval pads or tilting pads, etc.
3. Coupling Types
Rigid connection or elastic connection; tooth-type connection or membrane connection; number of connecting bolts; whether it has positioning sleeves, etc.
4. Gearbox Conditions
Such as gear types, layout; number of gear teeth, gear ratio, etc.
5. On-Site Fault Conditions
① Description of fault phenomena. For instance: how the fault was discovered, descriptions of fault manifestations from the operating personnel on duty, including relevant parameters shown on instruments; changes and developments in the fault at the time of diagnosis; conditions of various process parameters; lubrication status of the unit; piping conditions; temperature conditions at major measurement points.
② On-site auscultation, visual observations, tactile sensations, and even olfactory observations.
③ If conditions allow, shutdown and startup situations.
6. Vibration Data Collection
a) Types of sensors used in measuring instruments.
b) Vibration displacement, velocity, and acceleration values across the entire frequency range (preferably using simple vibration measurement instruments). Are the parameters single peak values or double peak values; are they effective values or others; it is best to use instruments with analog output indicators for observing whether there are fluctuations in vibration values.
c) Collected spectrums, preferably providing sampling frequency parameters, indicating whether it is amplitude spectrum, power spectrum, or inverse spectrum, etc.
d) Regarding shaft trajectory, indicate whether it is a simulated trajectory.
e) For phase information, indicate whether it is derived from FFT calculations.
f) Compare all data and spectrums, trajectories, phases, etc., before and after the fault (especially for gearboxes).
7. Special Inquiries
What kind of repair work was done on the machine or unit before the fault occurred, what parts were replaced, whether any machining was performed (which parts, which components), and who carried out the repair work, including reasons, processes, and outcomes of the repairs.
8. Pay attention to finding relevant equipment drawings and materials; if necessary, contact the manufacturer.
To the experts, scholars, and engineering technicians dedicated to the field of fault diagnosis!
Statement: This article is a thematic report by Xiao Ze, a member of the Vibration Forum, during a vibration salon event. The copyright belongs to the original author. Reproduction must indicate the source (Vibration Forum: vibunion.com or Vibration Home WeChat public account: vibunion).
