Condition Monitoring

Condition monitoring refers to the continuous monitoring of the condition of a machine, vehicle or entire industrial plant.

Condition monitoring uses sensors to continuously record (measurement) data, collect it in data storage devices (data loggers or PLCs) and analyse it using software. The sensors used measure physical values, such as rpm, temperature, speed, vibration and fill levels of operating and auxiliary materials. The individual measured values recorded by sensors are then digitised, processed and visualised. This provides an overview of the current condition of plant or machinery.

Continuous condition monitoring detects changes, malfunctions, defects and parts subject to wear and tear that have reached the end of their life cycle. It detects such changes before these can cause failures or damage to the plant or machinery being monitored. Components subject to wear may include damaged or worn roller or friction bearings, non-sealing valves, leaks or similar.

Continuous condition monitoring and comparison of actual and target conditions enable early detection of deviations in operating processes and the functioning of individual components.

Condition monitoring means maintenance is no longer tied to a fixed time or interval, but instead can be based on the actual condition of the plant or machinery.

Expensive downtimes can be significantly reduced by using condition monitoring, with maintenance and servicing tasks being carried out faster and only on components that really need it. A further feature of condition monitoring is increased safety for operators. Hazardous, "invisible" material fatigue, e.g. in tool holders, can be prevented by the continuous monitoring of the condition of plant or machinery.

Condition monitoring and predictive maintenance can mean very similar things, depending on the measurement set-up and the software. Predictive maintenance is not always necessary or even possible. Differences between condition monitoring and predictive maintenance lie in the capability of the software and the algorithms used.

Condition monitoring continuously records physical values such as temperature, sound, vibration, etc., stores these values and displays and partially evaluates them in software. Condition monitoring therefore primarily assists personnel in maintaining plant and machinery. Personnel get notified by software of limit value violations, or use their experience and knowledge to evaluate data to enable prompt maintenance or inspection of equipment. Analysis is therefore performed by personnel evaluating data and machine condition using the information provided by the condition monitoring system. This saves on costly manual inspections of system components.

Predictive maintenance uses the same sensor set-up on the machine as condition monitoring. The measurement data acquired is also the same. The difference lies mainly in the software and the algorithms used, also referred to as artificial intelligence or AI. Algorithms in the software enable independent assessment of the condition of plant and machinery. The software is therefore very capable of performing the function of experienced personnel. At best, the software can even detect correlations from multiple measured variables, correlations that have remained unknown until now due to their complexity. The predictive maintenance system automatically initiates maintenance of machines or shuts down plant components before a breakdown occurs.

Condition monitoring systems are also capable of independent switch operations and can respond according to pre-set limit values. However, condition monitoring cannot assess complex equipment conditions with multiple interrelated parameters, recognise patterns or make decisions based on these. Predictive maintenance requires developing a separate algorithm for each specific application and training it in a so-called "teach" phase on how to assess acquired data. This is not always possible and is associated with high levels of risk in certain areas.

Possible areas of application for condition monitoring in industry include the monitoring of:

• Rotating component speed (rpm)
• Torque (Nm) applied to rotating parts
• Vibrations and oscillations in bearings and shafts
• Electrical currents (A, mA) and electrical voltages (V, mV)
• Acoustic monitoring (ultrasonic testing, sonic testing)
• Linear acceleration (g-forces in x-, y- and z-directions)
• Temperatures
• Relative humidity
• Liquid levels
• Particle count monitoring in fluids such as hydraulic, engine and transmission oil
• Particle count in the ambient air (air filter quality)
• Gas analysis (CO2, CH4 etc.)

Condition monitoring in real conditions usually involves simultaneously acquiring and evaluating several of the above variables.

Condition monitoring requires both sensors and measuring devices as well as software to analyse and visually process the data. How the software works is explained later in this text.

Sensors detect reduced performance in electrical and mechanical components in plant and machinery:

• Vibration sensors detect imbalances, wear on bearings and misaligned shafts
• Pressure sensors detect flow turbulence and combustion dynamics
• Speed sensors are often used together with vibration sensors and are therefore capable of correlating vibrations with angular positions or rotational speeds
• Temperature sensors detect frictional heat
• Oil sensors detect contamination in oil and abrasions from gear cogs

Measurement devices acquire measurement values from sensors and convert these into measurement data. Measurement data acquisition in the professional and industrial sector is usually performed using data loggers, PLCs or small controllers rather than basic measuring devices.

Basic measurement devices are often unable to store or preprocess data efficiently. Data loggers or PLCs, on the other hand, enable large volumes of measurement data to be stored redundantly, preprocessed and transferred to condition monitoring software in a standardised data format.

Delphin Technology provides a wide range of professional and precision ICA systems for condition monitoring for use in a wide range of applications:

• LoggitoLab: Precision measurement technology for laboratories, test facilities in schools and universities, and for fault detection and analysis in plant and machinery.
• Loggito Logger: Universal data logger and small controller for any professional requirement: decentralised measurement networks, building monitoring, laboratory data acquisition, building monitoring, test stands, ...
• ProfiMessage: Modular PLC and acquisition and control of measured values for process data acquisition, fault value acquisition, trials, tests and product inspections.
• ProfiMessage D: Modular PLC and acquisition and control of measured values for process data acquisition, fault value acquisition and data automation for experiments and test stands.
• Expert Logger: Intelligent data logger and controller for industrial plants, laboratory data acquisition, environmental measurement technology, measurement data diagnosis, energy optimisation ...
• Expert Vibro: High-precision and high-speed data logger for acquiring transient signals and vibrations for shaft vibration monitoring and analysis and bearing fault diagnostics.
• Expert Transient: Data recorder/data logger for synchronous recording of transient and periodic processes. Optimised for troubleshooting in machinery, plant and equipment.

In a white paper, the UK-based TWI, an independent research and technology organisation, has summarised in three steps the basic process of setting up a condition monitoring system:

1. Install monitoring system

For plant or machinery to be capable of acquiring physical measurements, it must be equipped with the appropriate sensor technology. Many manufacturers already equip modern machines with the necessary sensors and interfaces. Older machines can be upgraded as part of a retrofit.

2. Establish baseline data

Condition monitoring is based on continuous comparison of a predetermined baseline of "ideal conditions" and actual measured values from plant or machinery. The baseline parameters need to be determined when the machine manufacturer has not already done so - most new machines already have their baseline set by the manufacturer. The required parameters, such as temperature, speed and vibration, are determined in several measurement runs and stored in the software as a baseline condition.

3. Start monitoring

After all baseline parameters have been established, condition monitoring can begin. Software now undertakes the processing of the data recorded by the sensors and automatically compares it with the stored parameters. Depending on settings, alarms can be triggered in the event of deviations, speeds can be reduced and any required replacement or worn parts can be automatically reordered. The software may also be capable of decision-making regarding required maintenance and servicing work.

Condition monitoring installation requires detailed advanced planning. The following five steps are essential when planning condition monitoring for your business:

1. Get the right machines

Machines need to be evaluated according to key figures and indicators to ensure that the right machines are selected for effective condition monitoring that provides your business with added value. Ideal here are strategic methods such as ECR (Equipment Criticality Ranking) or RCM (Reliability Centered Maintenance).

2. Get the right employees

Getting capable staff to carry out condition monitoring is key to its success. Many businesses make the mistake of entrusting the task of condition monitoring to their best mechanics, and not checking whether these also possess the skills needed to work with the new technology and IT.

3. Introduce condition monitoring training and education

Most staff who are to work with a new condition monitoring system receive insufficient, if any, training in how to use the system, interpret the data obtained or manage the output. Specific training and education are, however, essential for the success of the entire condition monitoring project.

4. Practice makes perfect

Every condition monitoring system should be extensively tested before going live. And not in a separate test environment, but in a 1:1 setup directly on the plant and machinery that is to be eventually monitored. This ensures consistency in the data obtained and avoids subsequent errors during operation.

5. Do not fear corrective action

After the condition monitoring system has been installed and the first data has been acquired, the next step is to take corrective action when the machine reports the first faults. What seems logical, however, is not always obvious for many businesses!

Condition monitoring is generally divided into four main groups. An extremely precise overview of a machine's condition can be gained by combining different groups. However, not every option is suitable or practical for every machine or plant.

1. Vibration and oscillation analysis

New vibrations and/or oscillations in rotating components, e.g. drive spindles in machine tools or bearings, are indicators of wear and defects.

2. Fluid analysis

Monitoring levels of operating fluids, auxiliary fluids and lubricants ensures safe operation of plant and machinery, and detecting leaks enables defective components to be identified and replaced.

3. Noise analysis

Every component that is set in rotation produces noise. Any faults in operation can be detected early on by using acoustic sensors.

4. Temperature analysis

Deviations in temperature can indicate a variety of defects. For example, increased temperatures may indicate wear and tear, while a sudden and significant drop in temperature may indicate a leak in a cooling circuit. Drive motors can also be monitored by thermal sensors.

Condition monitoring may be a very abstract term, but the data pathway from the point of origin to the point of analysis and evaluation is key to its understanding. A small example here will help you to understand the basic principles of condition monitoring. The temperature of an electric drive motor is being monitored.

1. The drive is equipped with a temperature sensor. The temperature sensor has been installed to monitor the temperature in the windings of the electric motor.
2. The sensor is also connected to a data logger. Data loggers such as the Loggito from Delphin Technology receive data from sensors, store it and forward it to condition monitoring software for further analysis.
3. During operation, the temperature of the motor suddenly increases by 12° Celsius. The temperature sensor reacts to this change by increasing its resistance value and thus "reports" the unusual temperature change.
4. The temperature change signal enters the data logger where it is preprocessed and stored.
5. The data logger is equipped with WLAN and connected to the company's cloud. The signal can therefore be transmitted almost in real time to the cloud and thus to the software.
6. The software automatically matches the digital signal to predefined limit values. Using data that has already been recorded in the past and supplemented by manufacturer information, an "overall picture" of temperatures in monitored the electric motor is created within the software.
7. The software then detects that the rise in temperature remains below the predefined top limit value, but is above predefined "normal values". The software also detects that the normal temperature range has already been exceeded several times in the past 14 days.
8. The software "decides" that a signal to switch off the motor (overload protection) is unnecessary. However, a message is issued that the temperature at the motor windings has again exceeded the normal range.
9. This message reaches the maintenance team as a push-up message on their mobile phones. They then decide to replace the component to prevent a total failure of the motor.

Condition monitoring used in technical systems has the potential to deliver significant cost savings and increased process reliability:

• Increased operational reliability through early detection of material fatigue
• Prevents sudden downtimes
• Maintenance can be specific and scheduled and therefore cost-efficientCost benefits from condition monitoring are particularly evident regarding maintenance.

Condition monitoring enables condition-based, intelligent and predictive maintenance. Continuously measuring physical states enables the actual remaining service life of technical components to be calculated. Downtimes can then be anticipated and maintenance planning can be carried out cost-effectively and resource-efficiently.

For many years, Delphin Technology has been actively partnering industry in the field of condition monitoring. Data loggers and measuring devices from Delphin Technology have been providing reliable measurement outputs, optimised condition monitoring, significantly increased process reliability and cost savings in the fields of mechanical engineering, energy technology and electrical engineering.

Do you want to know more about the cost benefits of condition monitoring? Then contact us now.

Offline condition monitoring is used for plant and machinery where continuous monitoring is not necessary or is not economically viable. Offline condition monitoring involves manual checking at set time intervals.

What does online condition monitoring mean?

Online condition monitoring is continuous, uninterrupted monitoring of plant and machinery. Condition monitoring here occurs without manual intervention or stoppages.

Delphin Technology is one of the world's leading specialists in vibration analysis and monitoring. The Expert Vibro measurement device enables transient signals and vibrations to be acquired even for major plants and projects. Expert Vibro has a compact design and can be coupled to multiple measurement devices to form a single unit. Expert Vibro is being used, for example, on test stands, in large construction machines and in the world's largest hydroelectric power plants to monitor turbine vibrations. Vortex tubes, plants and individual components can also be monitored reliably and continuously using Expert Vibro. The measurement device's capabilities are supplemented by ProfiSignal, condition monitoring software from Delphin Technology. The "Vibro" option turns ProfiSignal into a powerful tool for any vibration monitoring task.

We will be happy to give you a detailed presentation of our state-of-the-art condition monitoring products. Contact our expert team now and receive a non-binding consultation.