Synchronous peaks appear at the shaft’s fundamental rotation frequency and its harmonics (e.g., 2X, 3X). These peaks reflect the normal operation of rotating machinery and help identify regular operational patterns.

To do this, we need to learn the specifications of each type of machine. We must consider its geometric characteristics and equipment behavior – like the number of motor poles and rotation speed. Motor X will be our example, with the following specifications.

In order to collect the data and perform the analysis, a good option is to carry out equipment condition monitoring. There are two main methods: online and offline.

The most common and efficient monitoring technique is vibration analysis (VA). It’s often used to identify early signs of wear rotating equipment, as well as predict possible faults and failures. By monitoring vibration levels over time, engineers can see if a machine is starting to show signs of trouble. This allows them to take action before a failure occurs, preventing costly downtime and damage.

As we learn about vibration monitoring, we must also understand how this data contributes to the prediction of asset failures.

For vibration testing to be effective, we must accurately measure the vibration waves.There are 4 general ways to measure these waves:

Once the sensors have the data, the platform takes over to interpret it, generating insights and sending alerts. With access to real-time data, emergency repairs can be avoided and ensure cost savings.

Simple: by monitoring vibration levels over time, engineers can see if a machine is starting to show signs of trouble. This allows them to take action before a failure occurs, preventing costly downtime and damage.

The Work Orders tool within TracOS™ assigns the responsible parties and gathers data on equipment, tools, and materials. It also monitors the work order status, with notifications for each update – all from a single place.

Displacement is less sensitive to changes in the condition of a machine. It’s commonly used to measure the amplitude of vibration, which can be a sign of a loose or damaged component.

TRACTIAN installed 100 Smart Trac sensors on Ahlstrom Munksjö’s machines, which started collecting asset data remotely and in real-time.

The sensors quickly identified a vibration anomaly in two motors that had bearing failures and incorrect fastening, and sent an alert to the team. Once the insight was received, the team acted on it before something more serious could happen.

Descubra como o Mercado Livre de Energia pode transformar a gestão energética da sua indústria, oferecendo flexibilidade, redução de custos e eficiência operacional.

The best parameter to use depends on the specific application. Generally, you should measure all three parameters to get the most complete picture of the asset’s condition.

The use of vibration sensors to collect data is becoming more affordable and beneficial for businesses, particularly those that want to improve their maintenance practices.

These peaks occur at frequencies lower than the fundamental rotation frequency. They can signify issues like slipping belts, bearing cage defects, or fluid flow turbulence, which are critical for detecting subtle mechanical problems.

The frequency domain representation of a signal reveals the different frequencies that are present in the signal. This information can be used to identify different components of the signal, much like different harmonics in a machine vibration signal.

By reducing the frequency of failures, we can also reduce downtime, equipment replacement, and loss of time and resources. Our goal is to provide high-quality technology that optimizes maintenance routines for teams.

The method you choose will depend on the specific application. If you need to measure the frequency of vibration, then an accelerometer or velocity sensor would be a good choice. If you need to measure the displacement of vibration, then a displacement sensor would be a good choice.

The main goal of maintenance is simple: to improve performance and increase profits. In order to achieve these, teams must focus on tasks that increase asset reliability and availability – in turn reducing downtime.

After learning a machine’s baseline behavior, we can start to identify patterns in the vibration diagrams. These patterns characterize problems related to anomalies in the asset. With this information we can build a base table to identify variations in vibration.

Acceleration is the most sensitive parameter to changes in the condition of a machine. It’s often the best parameter to use for early detection of problems.

Smart Trac performs vibration checks and analyzes data over three axes: radial, horizontal and vertical. It collects data between 0 and 32kHz every 5 minutes, providing users with a full waveform and frequency spectrum.

Once set up, Smart Trac transforms vibration signals into electrical ones, automatically sending them to be recorded. The implementation is the fastest on the market – the plug and play system takes about 3 minutes to set up.

Spectral or spectrum analysis is a way of breaking down a signal into its component frequencies. To do this we use the Fourier transform (FFT) to convert frequencies from the time domain to the frequency domain.

Vibration analysis can uncover a range of mechanical issues, from misaligned components to structural weaknesses. It is particularly effective in verifying the proper installation of parts like bearings, ensuring they function correctly.

The application’s machine learning models function much like a neural network because they aren’t connected to the internet, but by 4G/LTE connectivity.

The TRACTIAN platform and sensors become a complete predictive system when combined with the CMMS TracOS™ software, which centralizes routines and automates processes.

Online monitoring employs sensors that combine Artificial Intelligence (AI) and Internet of Things (IoT) to collect data automatically and in real time. Because of this convenience, the presence of a maintenance professional is not required. In offline predictive strategies, manual sampling is performed and depends directly on the professionals.

In addition to preventing breakdowns, vibration analysis can also help you improve plant efficiency, reduce costs, and avoid supply chain issues. Identifying and addressing abnormal vibrations before they cause problems is the goal. By doing this you can improve the performance of your machines and reduce the need for unnecessary maintenance.

Maintainers verify if an unusual vibration pattern is related to a failure. Then, using spectrum analysis, we assess the anomalies, ensuring the machine’s health and proper functioning.

Grasping the nature of vibration signals is essential. Steady-state signals are continuous and repetitive, whereas transient signals occur due to specific events, providing critical diagnostic information.

The benefits of vibration analysis extend beyond cost savings and uptime improvement. By preventing breakdowns, you can also improve your company’s competitive edge and profits. This is because breakdowns can lead to lost sales, productivity, and customer satisfaction.

Vibration analysis can help you predict when maintenance will be required, so you can avoid costly breakdowns and extend asset lifespan.

Asynchronous peaks do not align with the shaft rotation frequency. They often indicate irregularities such as gearbox issues, pump cavitation, or periodic impacts, making them valuable for diagnosing unexpected problems.

Good news is: you can apply this technique in a wide array of assets, like motors, bearings, gearboxes, rotors and so on.

Then, the collected and recorded data is analyzed by a trained professional with the help of AI that assesses machine condition.

Descubra como aplicar a análise de causa raiz (RCA) para identificar e resolver problemas na indústria, aumentando a confiabilidade e a eficiência dos ativos.

Vibration analysis it’s often used to identify early signs of wear rotating equipment, as well as predict possible faults and failures.

Investing in vibration analysis and predictive maintenance technology is essential for anticipating issues and avoiding unexpected failures.

Cientista de Dados, engenheiro da computação pela USP. É especialista em AI, machine learning, deep learning, intelligent data analysis (analise inteligente de dados), data mining (mineração de dados), data compression (compressão de dados) e manutenção industrial. É Head de Dados e Sócio na TRACTIAN.

IoT sensors like the TRACTIAN Smart Trac are designed to be positioned at strategic points along an assets’ main axis line.

One of the most efficient ways to do this is by monitoring machine conditions. Condition monitoring lets us assess each asset, giving us crucial information about their behavior and current condition.

Ahlstrom Munksjö’s maintenance team acted swiftly to replace the bearing before it failed. This prevented unplanned downtime and the interruption of an entire production.

All machines emit a vibration signature, or specific vibration profile when operating. With continuous monitoring and analysis, anomalies can be detected in the vibrations of that equipment and its components.

They wanted to improve their condition-based maintenance (CBM) program to avoid equipment failures and production interruptions. They chose TRACTIAN to help them achieve this goal.