Electric Analysis

The Role and Importance of Electrical Signature Analysis in Predictive Maintenance

It is a well-known fact that unexpected downtimes of the machines in the industry and disruption of the production processes cost a fortune. Specific maintenance strategies have been implemented in the industry to prevent these downtimes. Some of these strategies are centered on the regular maintenance of the machine at certain periods. However, this kind of strategy reduces the efficiency of the production process by taking the machine into maintenance even if the machine is working in a healthy condition, or when there is no possible malfunction.

Condition monitoring-based predictive maintenance, which is one of the maintenance strategies developed for maximizing the efficiency to be obtained from the machine during the production process, has recently come to the fore with the increasing popularity of digitalization and sustainability topics. Condition monitoring enables one to come to conclusions about the current health status of the machine by taking certain data (vibration, current, voltage, temperature, etc.) from the machine and analyzing them. The point where predictive maintenance comes into play is to provide information about the malfunctions that may occur in the machine in the near future and their root causes by tracking and analyzing the data obtained and the results obtained with condition monitoring over time.

Condition Monitoring Techniques

The initial phase in condition monitoring is basic inspections. Small changes, such as abnormal heat or pressure, strange noises, excessive vibration, or a particular odor, are typically signs that something isn’t working properly. All techniques of condition monitoring are employed in a large variety of systems, from the most basic checks to cutting-edge tools. There are several condition monitoring techniques, and each is tailored to a particular machine component. When checking the state of various points of the machines, it is important to choose and apply the proper condition monitoring techniques according to the point to be applied since different forms of failure can be detected by different condition monitoring techniques.

In addition to electrical signature analysis, there are condition monitoring methods such as oil analysis, ultrasound analysis, infrared thermography, and acoustic emission analysis. Each of them has a different capacity to detect malfunctions that may occur in the machine a while ago. For example, oil analysis and ultrasound techniques can predict possible malfunctions earlier than vibration and electrical signature analysis methods, even when the machine is in good condition and shows no visible signs of malfunction.

Potential Failure – Functional Failure Curve

Fig. 1 Potential Failure – Functional Failure Curve

As the fault progresses in the machine, the repairment of the fault becomes costlier and requires more time. This means more unexpected downtime. On the other hand, the condition monitoring method to be applied should be an economical solution. Since the measurement devices and application required for condition monitoring with the mentioned oil and ultrasound analysis methods are costly, their application may not be possible everywhere. Besides, each method is specialized in finding certain failure modes. Therefore, vibration analysis and electrical signature analysis methods, which have a wider scope and application, have found more place in predictive maintenance applications.

The Aspects of Electrical Signature Analysis that Differ from the Other Methods

41 percent of asynchronous motor failures are caused by the bearings, 37 percent by the stators, and 10 percent by the rotors. Although bearing failures can be detected more clearly with vibration analysis, they can also be detected by electrical signature analysis as well. On the other hand, in asynchronous motors, especially in stator and rotor faults, the electrical signature analysis provides a very clear detection and is a very powerful and economical solution compared to other methods.

Mechanical misalignment or unbalance, and air gap misalignment can be detected by both electrical signature analysis and vibration analysis. Failures caused by motor supply and insulation problems can be detected by electrical signature analysis. Electrical signature analysis is much more powerful than other detection methods in detecting broken rotor bars and stator problems.

Since each condition monitoring method can analyze the health status of different parts of the machine, and detect possible faults that may occur in these parts with their root causes, increasing the application of these methods enables the detection of faults in a wider range. Since electrical signature analysis is easy to implement, it can be easily preferred instead of or alongside other condition monitoring methods.

Electrical Signature Analysis

Fig. 2 Motor Control Center (MCC)

Unlike other condition monitoring methods, electrical signature analysis is not dependent on machine operating conditions to collect data. Because the sensors to collect the data are not applied directly on the machine, but in the cells called the motor control center (MCC) or the panel where the electrical connections of the machine are made. Here, the current and voltage values of the motor are measured by attaching current and voltage sensors to the cables feeding the motor. Since the ambient conditions in the motor control center are generally the same, measurements can be taken continuously, and the data can be analyzed uninterruptedly.

Since current and voltage data are very clear information, these data can be easily obtained accurately by applying economic solutions with basic measurement methods. Making sense of data is also easier compared to other condition monitoring methods. By analyzing the data in the frequency space, the presence of the fault can be detected by the formation of sideband components with equal frequency intervals to be formed around a fundamental frequency (motor driving frequency). The harmonics of the sidebands can also be observed. The main part where electrical signature analysis differs from vibration analysis is that harmonic frequency components in multiples of the fundamental frequency are observed in the vibration analysis, while electrical signature analysis observes sideband and sometimes sideband harmonics.

Including Electrical Signature Analysis in Condition Monitoring Strategies

In addition to collecting and analyzing vibration data with vibration and temperature sensors Infinity and Wired, Sensemore also collects and analyzes electrical data with current/voltage sensors and IoT data collection device Duck. There are 8 different channels for Duck to collect data from different sensors simultaneously and transfer it to the cloud, and 3-phase current and 3-phase voltage information in motors can be collected simultaneously using 6 of these channels. In this way, methods such as motor current signature analysis (MCSA), voltage signature analysis (VSA), and instantaneous power signature analysis (IPSA), among the techniques included in the electrical signature analysis, can be applied.

Current and voltage data are taken from the phase cables feeding the motor via sensors and transmitted to Duck. Duck transfers this data wirelessly to the cloud. Data can be viewed and analyzed via the cloud application. Possible malfunctions that may occur in the machine are notified to the users in advance and it is ensured that the users can take action before unexpected stoppings occur.

Electrical Signature Analysis

Fig. 3 Application of Electrical Signature Analysis

References:

  • World Economic Forum, Analysis: Global CO2 emissions from fossil fuels hits record high in 2022.
    United States Environmental Protection Agency, Sources of Greenhouse Gas Emissions


Fault Diagnostic Technique

Fault Diagnostic Technique Using AI-Assisted Machine Mode Similarity Analysis

Artificial intelligence applications are widely used for fault diagnosis by monitoring vibration data. In order for artificial intelligence applications to provide the most efficient results, they need to be trained using data obtained by regular monitoring of machines for long periods and evaluated by experts. As a result of this training, the operating characteristics of the machinery will be recognized and the diagnosis of the potential faults will be possible. However, even during the training process, machine failures must be detected successfully. In such cases, where there is not enough data readily available yet or if it is difficult to obtain working characteristics from experts, a diagnostic technique based on machine mode similarity can be used.

Raw vibration data from the sensors can be converted into meaningful features and using these features, “machine modes” that summarize the operating cycle of a machine can be found. Machine modes indicate important changes that a machine may encounter during its lifetime, such as machine downtime, minor changes in operating conditions, and malfunctions. The extracted modes indicate similar trends for similar machine types used in industry. Therefore, the operating characteristic of a machine for which there is insufficient information or measurements, can be determined using the modes of another machine whose operating characteristics are well known. A process, which could take a lot of time and effort under normal conditions, becomes easy, fast and understandable with mode similarity analysis.

Mode Similarity Analysis

For the mode similarity application, determination of a “donor” machine is the first step. A donor machine is a machine which is well recognized by our artificial intelligence algorithm and it provides the machine mode information to the system. The raw vibration data coming from the donor machine are analyzed with various techniques in time and frequency domains and decomposed into their features. These features are evaluated by the machine learning algorithm and their respective signals are separated into machine modes. The resulting modes are examined by reliability engineers who are experts in the field, in order to determine which operating conditions or fault types they belong to. Simultaneously, the data of the “acceptor” machine, are processed and its features are extracted. Acceptor machine is the machine whose status is desired to be known and to which the machine modes will be transferred. An acceptor machine should have a similar drivetrain to the donor machine. Finally, the failure modes of the acceptor machine can be determined by analyzing the similarity between the defective modes of the donor and available measurements of the acceptor machine. In this application, a real machine can be used as a donor as well as a readily available, comprehensive mode library.

Mode Similarity Analysis Flow

Fig. 1 Mode Similarity Analysis Flow

Example Application

We can demonstrate this technique more clearly with an application. First of all, let’s consider a centrifugal pump driven by a 250 kW electric motor as the donor machine, whose operating characteristics are fully defined in our mode pool. As the acceptor machine, a fan driven by an electric motor with a capacity of 200 kW will be considered. Although the basic features of these two machines are very different, if the raw vibration data is analyzed correctly, it will be possible to transfer the machine modes by establishing a correlation between them. Both of the selected machines have variable rotational speeds. Therefore, the resonance situation, which has a critical importance in machine health, has been chosen as the type of failure to be determined in the scope of the example. However, a similar application can be applied for many other failure root causes.

The 3-axis vibration data as well as the measurement specific RPM (revolutions per minute) values were collected from the donor machine. Meaningful features were extracted from the raw data and were fed into the machine learning algorithm. The algorithm outputs the measurements divided into machine modes. The mode that shows the resonance characteristic and the measurements included in this mode were marked for comparison with the acceptor machine.

Resonance Mode

Fig. 2 3 Axis VRMS Data with Respect to Measurement Index for the Donor Machine, Separated Into Machine Modes Including the Resonance Mode

For the acceptor machine, the measurements were collected in the same way and their features were extracted. Then, as a final operation, the measurements of the mode marked as resonance failure in the donor machine and all the measurements of the acceptor machine were fed into the similarity algorithm. The resulting similarity score for each measurement indicates how close the measurement is to the resonance mode. The measurements with the highest score should then be re-examined by the reliability engineer team to fully confirm the failure mode.

In this application, the measurement number 341 has the highest similarity score and is confirmed to show the resonance characteristic for the acceptor machine.

Conclusion

Within the scope of predictive maintenance, even if a machine is still in the recognition process, its operating and failure modes can be determined by making mode similarity comparisons. In this way, it is possible to have an idea on the instant status of many machines without waiting for the initial training and estimation phases, and thus quicker reactions can be given in case of emergencies.

References:

  • Şerifoğlu, M. O., Gencer, F. B., Aktaş & A. Ö., Ulusoy, A. E. (2022). Makine Modu Benzerliğini Kullanarak Titreşim Tabanlı Rezonans Teşhisi. Uluslararası Katılımlı Bakım Teknolojileri Kongresi ve Sergisi 20-22 Ekim 2022 Denizli. ISBN: 978-605-01-1546-8


Pharmaceutical Industry

Pharmaceutical Industry and Predictive Maintenance Applications

Pharmaceutical Industry is one of the largest fields of production in the world. With an incredibly demanding clientele, challenging development and production processes and high volumes of production, Pharmaceutical Industry is forced to be a pioneer to research and implement new technologies into their production lines. These innovations are vital to keep up with the constantly increasing needs of the world. Efficiency, quality, sustainability and safety are some of the main traits focused on by the innovators in the field of pharmaceuticals. In order to achieve improvements in these traits, production lines of pharmaceutical companies must be in correlation with innovative technologies regarding every aspect of the production. Maintenance is one the fields that require attention to ensure production lines are efficient, sustainable, safe and able to produce high quality products. In this blog post the main focus will be on predictive maintenance applications that are fitting for the needs of the pharmaceutical industry.

Pharmaceutical Industry and Predictive Maintenance Applications

A large variety of machines are used in the pharmaceutical industry. Processing equipment include agitators, centrifuges, blowers, capsule equipment, chillers, pulverizers etc. Post production equipment include CAM Blister Packing Machines, Bottling Lines, Counters, Sealers etc. All of these equipment works with small size products in a highly complicated, fast and precise manner. That’s why components used in these machinery are compact, precisely machined and purpose built. If any one of the critical components in these complicated machines were to fail, quality of the product can be heavily affected and all of the production line might face an unforeseen stoppage. In order to prevent such stoppages and malfunctions maintenance teams need to implement predictive maintenance technologies. Main elements providing and conveying the movement to the machinery are rolling elements, gearboxes, servo and induction motors. Predictive maintenance applications can be performed through many different applications such as thermal monitoring, lubrication analysis, ultrasonic inspection, vibration analysis, electrical analysis etc. Among these different analysis approaches vibration and electrical analyses are the most viable ones due to the ease of application and affordability.

Pharmaceutical Industry and Predictive Maintenance Applications

Electrical analysis is one of the predictive maintenance applications fitting to the needs of the pharmaceutical industry. Especially on the components such as servo and induction motors used for the generation of precise and continuous movement to the production line. IoT based data acquisition devices paired with current transformers and voltage dividers can provide precise data about the energy consumption and efficiency of the equipment. These measurements can be analyzed to diagnose many different modes of malfunctions such as stator, rotor and connector problems and misalignment, unbalance and bearing failures before they result in a serious failure.

Vibration analysis is widely used in industry for the predictive maintenance of rotating machinery. Vibration data is usually collected through accelerometer instruments. This data can indicate many different modes of malfunction if collected and analyzed correctly. Many generic failure types such as mechanical looseness, unbalance, bearing failure, cavitation, gear failure, lubrication fault, belt and pulley fault etc. can be detected through vibration analysis. Sensemore AI powered predictive maintenance applications can identify anomalies indicating the root causes of malfunctions months prior to the occurrence of the malfunction.

Foreseeing failures, estimating remaining useful life and eliminating catastrophic malfunctions through early detection and corrective actions is the key to create an industry with higher efficiency, quality and sustainability.


Sustainability

Contributing to a Sustainable Future by Improving the Maintenance Strategies

As a result of the gradual depletion of the resources in the world and the problems such as global warming, excessive concreting, air and water pollution, and loss of biodiversity; it is clear that it is becoming more difficult for all living things to coexist on Earth and for the Earth to remain in a certain state of balance day by day. As sustainability can be used in a social or economic context, since it concerns all living things, its use in daily life focuses on the sustainability of environmental conditions in a balance and actions that can be taken for a green future.

Since there cannot be a direct transition to a completely sustainable lifestyle in the current situation, some actions are taken to get used to this transition. Some of these are reducing food waste, reducing birth rates and population growth, opting for plant based products and applying it in nutrition, promoting green technologies, reducing fossil fuel consumption, and promoting the use of renewable energy.

Industrial Carbon Emissions

In order to limit global warming to 1.5 °C, carbon emissions in the world should decrease by 45% by 2030, and “net-zero” should be reached by 2050, that is, carbon emissions caused by humans should be zeroed. So far, about 120 countries and 1000 companies have set a “net-zero” target [1].

Generation, transmission and distribution of electrical energy, and industrial activities account for the majority of carbon emissions. A lot of resources are consumed for these activities and these activities are mostly carried out with non-renewable resources such as fossil fuels. Sustainability can come to the fore by reducing the use of such non-renewable resources.

Industrial carbon emissions can be studied directly and indirectly. Direct carbon emissions include the fossil fuel consumed to power the machines and the warming it causes. If natural gas or oil is used to energize such machines, leakages in their transmission also significantly affect carbon emissions. Indirect carbon emissions arise from the use of non-renewable resources (fossil fuel, oil, natural gas) for the production of electrical energy to be used for other purposes.

Factories

Coal, natural gas, oil, and fossil fuels are consumed for the generation of electrical energy. Although the share of coal in electrical energy generation in the USA in 2020 is 20%, the share of used coal in total carbon emissions has been determined as 54% [2]. The use of renewable energy sources for energy production will reduce carbon emissions.

In addition to the generation of electrical energy, when calculating carbon emissions from electrical energy, the carbon emissions that will be caused by the way it is used when it meets the end user should also be taken into account. The industry has a share of approximately 30% in terms of carbon emissions caused by the use of distributed electricity, and when this is combined with industrial carbon emissions, the carbon emissions caused by the industry are at the forefront.

Enabling the Sustainability

Reducing industrial carbon emissions will have a significant impact on sustainability. In order to do this, it is necessary to reduce the consumption of resources such as oil by performing regular maintenance of the machines and ensuring that the energy consumption does not exceed the expected values.

With effective maintenance strategies, maintenance costs can be reduced by up to 40% and energy consumption by 10%. In addition, with effective maintenance strategies, the risk of unexpected downtimes in production is reduced and the cost of a product is somewhat reduced. This contributes to the reduction of carbon emissions.

Machine health monitoring to reduce carbon emissions can be done most effectively with predictive maintenance. With certain condition monitoring methods, data is received from the machine and these data are analyzed. By following the changes, inferences are made about a possible failure in advance. Users are informed about possible malfunctions, the machines are examined during planned downtimes, the machines in need of maintenance are detected and asset loss is prevented.

Mechanical and electrical faults cause more energy to be drawn from the mains in the machines, but this energy is transformed into vibration and heat instead of being used in the machine. Regularly maintained and healthy machines consume less electricity compared to their malfunctioning conditions and play a role in reducing carbon emissions. In addition to that, equipment that is regularly maintained has a lower risk of releasing substances that are harmful to the environment.

Monitoring

Regular maintenance will increase the life of the machine. This means that the components in the machine will last longer. The long lifespan of the machine elements indicates that the need for component replacement will be less, and it contributes to a sustainable future as there will be no carbon emissions for the production of these machine components.

Thanks to strong maintenance strategies and predictive maintenance applications, the machine’s health can be improved and the frequency of maintenance required by the machine can be reduced, the source of the failure can be determined in advance, thus reducing the time period for maintenance personnel to spare for maintenance, thus saving time for different tasks. In terms of sustainability, carbon emissions are reduced by saving materials and labor spent on maintenance, besides extending the remaining useful life of the machine.

Implementing an Efficient Maintenance Strategy

For the above-mentioned reasons, maintenance strategies should be preferred in order to achieve sustainability targets, where the machines will be maintained in the least amount and time and the unplanned downtime will be the least. In this sense, the reactive maintenance strategy, which recommends taking the machine into maintenance when there is a malfunction, will be quite inefficient. Planned maintenance is inefficient because it does not give an idea about the malfunction and is carried out independently of whether there is a malfunction or not, as it can cause loss of healthy operating time, as well as cause loss of long time and resources until the source of the malfunction is found in the event of a real malfunction. The most effective method here would be condition monitoring based predictive maintenance application.

Condition monitoring provides information about the current health status of the machine by obtaining certain data (temperature, vibration, current/voltage, etc.) from the machine and analyzing these data. Predictive maintenance is a maintenance strategy that can predict future failures using the change in machine data and health.

Sensemore products

Sensemore collects data from the machines with the software and hardware solutions and the services it offers to its customers, and analyzes this data in the cloud application interface. Vibration, temperature, current, and voltage data can be analyzed with the hardware and the data obtained from all kinds of analog sensors can be displayed in the cloud application. Users are informed with root-cause analysis of possible failures and contribute to sustainability goals by preventing unplanned downtime.

Thanks to the energy monitoring application that Sensemore offers to its customers with the IoT data collection device, voltage sensors, and current sensors, it allows for determining the amount of energy that the machines consume and reducing carbon emissions by taking action on the machines that consume more energy than normal.

References:

  • World Economic Forum, Analysis: Global CO2 emissions from fossil fuels hits record high in 2022.
    United States Environmental Protection Agency, Sources of Greenhouse Gas Emissions


Maintenance Culture

History and Future of Maintenance Culture

There are many different explanations of what maintenance is in the technical field. It is a wide concept that contains tests, measurements, replacements, adjustments, repairs etc.that aim to increase the overall health of a system. Even though this broad explanation of maintenance is true for all of our history, maintenance culture is filled with different approaches and procedures for different parts of humankind’s technological timeline. In this blog post, the main focus will be these different maintenance cultures throughout the history and the future of maintenance.

For the majority of our existence, humanity relied on a maintenance approach called Corrective Maintenance. This approach is the earliest one among the maintenance concepts. Malfunctions are addressed and systems are maintained when a failure comes to existence. This approach is solely reactive since systems generated before the industrial revolution contained no condition monitoring or failure analysis process. Due to the smaller volumes of production and less demand from the market, this corrective maintenance approach did not create much of a problem for the pre-industrial revolution era of production.

After the Industrial Revolution maintenance turned into a higher importance issue for plants around the world. Especially widely used boilers started to become a threat for workers due to the commonly occurring failures resulting in explosions. Industry started to implement periodic evaluations on these boiler equipment in order to ensure safety in the work environment. This new method of ensuring the health of the system marks the emergence of a new maintenance approach called Preventive Maintenance. Preventive Maintenance holds the aim of eliminating the possibility of malfunction through periodic maintenance processes at heart. With the increasing demand from the market and growing production volumes this approach started to spread from boilers to a wide variety of machinery in various plants.

Throughout human history wars have always been a driving force behind innovative technologies. After World War 2 industry started to focus more on reliability and availability of systems since it would grant the machinery the vital ability of being able to operate in short notices successfully. This resulted in the birth of a new maintenance approach called Proactive Maintenance. Proactive Maintenance relies on fault analysis and condition monitoring. Fault analysis methods enable users of a system to evaluate a failure alongside its root causes, conduct criticality analysis and eliminate conditions resulting in a malfunction. Condition monitoring systems inform users on vitals of the system through sensors connected to critical equipment. The constant inspection and flow of information enables the operators of the system to observe any differences in the condition of the machinery and to take action before a malfunction occurs.

Factory Worker

Nowadays with the increased availability of various sensors and the emergence of the Industry 4.0 concept, condition monitoring technologies became more popular in production sites. The ability to acquire large amounts of data from the machinery facilitated the creation of a new maintenance approach called Predictive Maintenance. Predictive Maintenance utilizes large amounts of data gathered through condition monitoring applications, and conducts analysis to understand the mode of malfunction and creates models to predict the root cause and the remaining useful life of machinery.

Future of maintenance holds infinite possibilities with the constant developments in the machine learning field and AI technology. Solutions such as Sensemore products use AI models to perform these analyses in an automated manner and predict the root causes of malfunctions before they occur. Using AI algorithms to go through large quantities of data gathered from systems, continuous condition monitoring through precise sensors and conducting root cause analysis is the future of maintenance culture.