28.5 Practical Insights and Case Study

Back

There are currently many companies in the industry that deal with machinery vibrations. Broadly,

these companies can be classified into those that produce vibration-related testing products (e.g.,

FFT analyzers [Steinberg, 2000], vibration data collectors, and end-of-line production test

equipment), and those that provide solutions in resolving noise- and vibration-related problems

for different industries (e.g., automotive, aerospace, manufacturing, and engineering companies).

For instance, if one of the machines in operation on the shop-floor is experiencing unexplained

machinery noise or a high level of product failure, vibration analysis (Farrar et al., 2001) may

provide some answers.

The case study explores the feasibility of remote vibration monitoring and control of a number of

machines or mechanical systems located at different sites. For the sake of generality, “machines” is

used as a general term to represent vibrating machines, mechanical systems, and other systems that

exhibit vibration behavior. There are strong motivations for such setups. Since the middle of the

1960s, the concept of the distributed system has been widely adopted in the process industries, such

FIGURE 28.39 (a) Vibration signature corresponding to the sinusoidal input, with a standardized amplitude of 1 V

and frequency of 5 Hz; (b) spectrum of the machine corresponding to the sinusoidal input after fault occurs.

SUMMARY

In this section, the development of a low-cost approach towards real-time monitoring and analysis

of machine vibration is described. A real-time analyzer, based on a fuzzy fusion technique, is used

to monitor continuously and compare the actual vibration pattern against a vibration signature.

This intelligent knowledge-based real-time analyzer is able to detect excessive vibration conditions

much sooner than a resulting fault could be detected by an operator. Subsequently, appropriate

actions can be taken. This approach may be implemented independently of the control system and

as such can be applied to existing equipment without modification of the normal mode of

operation. Experimental and simulation results are provided to illustrate the effectiveness of this

real-time vibration analyzer.

Vibration Suppression and Monitoring in Precision Motion Systems 28-29

© 2005 by Taylor & Francis Group, LLC

as the chemical industry, manufacturing industry, food processing industry, primary industry, and

many others. These distributed systems are also characterized by smart devices. As the costs of

manufacturing, that is, labor costs, the cost of raw materials, fixed costs, rental costs, and other

production costs, are a major concern for the location of plants, many processes are currently

widely distributed geometrically. The layout of an entire plant can be rather extensive, spreading

across continents in certain cases.

The extensive distribution of an entire plant requires close coordination and synchronization of the

distributed operations, as well as an efficient remote monitoring and control facility in place. The

utilization of the Internet for this monitoring and control purpose perfectly complements the trend of

distributed intelligence in the field controllers and devices on the shop-floor. As a result, superior control

decisions can be made with these readily available resources and information, that is, historical data,

knowledge databases, and so on. Internet working has become essential for plant automation. A case

study which implements remote vibration monitoring and control of various machines will be illustrated

in this section.

This vibration monitoring and control system possesses the structure as shown in Figure 28.40. It is

able to monitor continuously the real-time health conditions of multiple machines at different sites,

connected via the Internet. The system operates in two modes: the learning mode and the monitoring

mode, as mentioned earlier. In the learning mode, vibration signatures, representative of the health of the

machines to be monitored, are stored in the supervisory controller. Accelerometers mounted on the

machine directly provide measurements of the vibration signals. In the monitoring mode, real-time

vibration signals are streamed from the front-end controllers to the supervisory controller. Based on

these real-time vibration signals from the various machines, the supervisory controller is able to generate

decisions on the health condition of the machines, taking into consideration various criteria based on a

fuzzy fusion technique, as mentioned earlier. Alarms will be activated when the health condition of the

machines falls below an acceptable threshold level. Subsequently, rectification actions, to provide a

warning or automatic corrective action (e.g., changing the operating conditions of the machine,

FIGURE 28.40 Remote vibration monitoring and control system.

28-30 Vibration and Shock Handbook

© 2005 by Taylor & Francis Group, LLC

modifying the parameters of the controller, or shutting down the machine), may be invoked before

extensive damage is done to the machines.

Referring to the architecture of the vibration monitoring and control system as shown in Figure 28.40,

a data transfer and communication protocol needs to be adopted for such an application. There are

several ways of connecting the supervisory controller and the remote machines. One of the possible ways

is to apply DataSocket technology; yet another is to use low-level protocols such as TCP/IP and User

Datagram Protocol. As can be seen, this vibration monitoring and control system may be applied to

existing setups at the shop-floor level, without much modification necessary. The front-end field devices

(i.e., actuators, sensors, and other I/O modules) are linked up via the field-level data highway. Two

popular types of field-level data highway are fieldbus and twisted pair. Access security is a main issue of

such remote applications. There is a need to prevent unauthorized users from accessing and modifying

the system so as to maintain the integrity and proper functioning of the system. This can be achieved by

FIGURE 28.41 Learning mode — snapshot of the control panel at the supervisory controller end.

Vibration Suppression and Monitoring in Precision Motion Systems 28-31

© 2005 by Taylor & Francis Group, LLC

imposing some access authentication scheme. One commonly used basic access authentication scheme is

the challenge – response authentication mechanism whereby the operator or user must provide some user

logon information (e.g., user identification and a password) before accessing the supervisory controller

terminal.

The control panel at the supervisory controller end (Figure 28.40) is as shown in Figure 28.41. At the

supervisory controller end, the operator is able to monitor and control the health condition of various

machines at different remote sites. As shown here, two machines (i.e., machine A and machine B) are

being monitored. The learning mode is first initiated to obtain the vibration signatures of the two

machines in the normal operational condition. The vibration signature of machine A resembles a chirplike

signal input whereas that of machine B resembles a sinusoidal-like signal input. At the end of

the learning mode, a message is shown in the message box. Figure 28.42 shows a snapshot of the

control panel in the monitoring mode. Here, the two machines are in good health, as the health indices

FIGURE 28.42 Monitoring mode — snapshot of the control panel at the supervisory controller end before any fault

is invoked on machine A and machine B.

28-32 Vibration and Shock Handbook

© 2005 by Taylor & Francis Group, LLC

(i.e., current values are at 0.9) of the two machines are above the threshold value of 0.6. To simulate a

fault arising in machine A, the input signal to machine A is changed to a sinusoidal input. This causes a

simultaneous change in the vibration pattern of Machine A. A snapshot of the control panel at the

supervisory controller end is as shown in Figure 28.43 after the fault is invoked to machine A. As can be

clearly seen, the vibration pattern of machine A has changed. The message box displays that the health

index of machine A (current value is at 0.49) has fallen below the threshold value of 0.6. To signal the

deterioration of the health condition of machine A, the alarm LED is lit up. The health condition of

machine B remains in the satisfactory region. To simulate a fault arising in machine B, the amplitude of

the input signal to Machine B is increased slightly, by 20%. Consequently, this will also cause

simultaneous change in the vibration pattern of machine B. A snapshot of the control panel at the

supervisory controller end, after the fault is invoked to machine B, is as shown in Figure 28.44.

FIGURE 28.43 Monitoring mode — snapshot of the control panel at the supervisory controller end when a fault is

invoked on machine A (at t ¼ 3 sec).

Vibration Suppression and Monitoring in Precision Motion Systems 28-33

© 2005 by Taylor & Francis Group, LLC

The vibration pattern of machine B has changed. The message box displays the health index of machine B

(current value is at 0.53) has fallen below the threshold value of 0.6. In the same way, the deterioration of

the health condition of machine B is highlighted by activating the alarm LED.

FIGURE 28.44 Monitoring mode — snapshot of the control panel at the supervisory controller end when a fault is

invoked on machine B (at t ¼ 5 sec).

SUMMARY

This section gives the reader practical insights into vibration monitoring and control applications

in the industry. A case study is provided to explore the feasibility of remote vibration monitoring

and control of a number of machines or mechanical systems located at different sites.

28-34 Vibration and Shock Handbook

© 2005 by Taylor & Francis Group, LLC

Навигация