Condition Monitoring:

Understanding the Options

In previous articles, we discussed the value of predictive maintenance in its effectiveness and the tangible benefits it provides to a company or organization. While other approaches, such as preventive or reactive maintenance, have use in specific situations, predictive maintenance can be successfully applied to nearly every piece of critical machinery or equipment.

The central tenet of predictive maintenance is the detection and resolution of failures before they occur through observation of the actual conditions of the machinery.  Machines can fail in a variety of ways, from the wear of a part, to environmental damage, to insufficient electrical power supply.  Other maintenance strategies may either go too far – preventive maintenance, which can perform upkeep tasks far too frequently and result in unnecessary expenditures – or not far enough – reactive maintenance, which involves not repairing equipment until it breaks, and usually causes costly stretches of unplanned downtime.  Predictive maintenance, in contrast, is “just right” – it focuses on observing and collecting data about the performance of the machine, understanding trends in the data that indicate failure may be imminent, and scheduling and performing maintenance only when it is necessary yet convenient to do so.

An important element of the majority of successful predictive maintenance strategies is a tactic known as condition monitoring.  Condition monitoring involves the application of specific hardware, software, or a combination of the two to gather and report data about the performance of various machines.  The technologies utilized in condition monitoring focus on the areas in which different types of failure are most likely to occur, creating a clearer picture of a machine’s health and productivity than any competing maintenance strategy.

The following guide outlines the different types of common failures and lists the technologies which have been developed to address each one.  It also discusses the benefits of installing condition monitoring hardware and why the Tactix™ system, developed by ProAxion®, is an excellent option due to its ease-of-use, cost and payback history.

Identifying Common Failures

In regards to maintenance, the term failure is defined as “the inability of a machine or piece of equipment to fulfill any one of its intended functions.” Typically, failure can be divided into three distinct categories based on the source of the problem:  mechanical, electrical, or operational.

Mechanical failures, the most common type, includes any failure which occurs due to wear, tear or other damage caused to mechanical parts.  Examples of mechanical failures include a motor slowing down, a shaft becoming misaligned, or a bearing wearing down due to age and extensive use.  Breaks, cracks and other physical damages are all categorized as mechanical failures.

Electrical failures, as the name implies, deal with any problems regarding the power source, typically electrical in nature, which provides energy to the equipment or machine. A blackout, or total loss of power, is the most frequently observed type of electrical failure. Other examples include an insufficient amount of energy being transmitted to a machine, causing it to slow down, or a circuit not being properly closed.  A failure can sometimes include both mechanical and electrical elements, such as a power outage caused by snapped wires or a torn cord.

Lastly, operational failure refers to failures that are caused by humans who operate the machines.  A common nickname for operator failure is “human error”.  There are an uncountable number of different ways in which operator failure can occur, and it is therefore sometimes considered the most difficult type of problem to predict or prevent.  Occasionally, an operator failure can even lead to a mechanical failure, such as a worn-down bearing caused by an operator forgetting to regularly clean the machine.  However, the risk of operator failure can be minimized by ensuring that all employees are properly trained in both the proper, safe use of the machines and in the performance of upkeep tasks such as cleaning and the replacement of lubrication oils.

Failure Modes: The What, Where, Why and How of Failure

In the world of machines, very few failures are truly random.  Certainly, “acts of fate” which cannot be prevented or planned for, such as natural disasters, do sometimes occur, but these are very much the exception rather than the rule. Most failures can be accurately predicted using a combination of historical data, averages and statistics regarding the machine and up-to-date data gathered by a combination of condition monitoring devices and, sometimes, inspection by human professionals.

In order to develop a successful condition monitoring device, companies such as ProAxion® look at the failure modes, or the most common causes and results of failures within each individual machine. When performing a failure mode analysis, a number of key questions must be asked.

  • Where, or in what part, of the machine does the failure occur?
  • What is the nature of the failure (broken bearing, power outage, etc.) and what are its effects on the machine’s performance (slow movement, creation of damaged products, total shutdown, etc.)?
  • Why did the failure occur (age, environmental damage, contamination, etc.)?
  • How can the failure be fixed, and how can it be prevented from occurring in the future?

Once these questions have been asked, at-risk parts and existing bugs or defects can be noted.  If it is known where in the machine a failure occurs, a condition monitoring device can be created which specifically targets that part and successfully records its performance over time.

Examples of Common Failures and Failure Modes

The majority of machines are made up of many small parts working together to create a fluid, efficient whole. Some of those small parts are significantly more likely to wear down, become damaged and fail than others.  It is those parts that the most effective condition monitoring hardware and software tend to target.

Rotating parts such as fans, motors, engines, and certain types of bearing, are common sources of failure.  Because these parts are constantly or usually in motion, and their continuing motion is necessary to the successful function of the machine, anything that causes them to slow down or stop can easily result in a failure.  For example, dust and grit can become caught in a fan, bearings can wear down from rubbing against each other, motors can slow down if they are not receiving enough power – the possible failures are numerous and may, at first glance, seem overwhelming.

Fortunately, there are a number of ways in which rotating parts can be observed in order to monitor their performance.  The most commonly used method is observing vibration, which can change in frequency or intensity as parts degrade or approach failure.  In addition, sound (which is a subset of vibration) can be observed, as the failure of a rotating part has often been known to cause “unnatural” sound such as high-pitched whining or loud grating noises.  Additionally, in certain rotating parts such as pumps or motors, pressure levels can be monitored to look for unnaturally high or low values.

Lubricating oils are another machine component known for causing or being the source of failure.  Lubricating oils can damage the machine if they are not replaced frequently enough – the levels can become too low (or too high through operational error) or the oils can degrade in quality to the point where they are no longer performing their intended function as a lubricant.  Additionally, these oils can easily become contaminated by dirt, dust, grit or particles caused by the wearing down of internal parts.  Condition monitoring devices have been measured which use a variety of methods to observe the quality of lubricating oils and provide alerts when actions such as checking the level or replacing the oil should be taken.

Similar to lubricating oils, any other liquid or fluid component within a machine can drain, leak, dry up, or fail as a result of contamination.  Condition monitoring measures either the quality or quantity of the fluid as discussed above or its pressure, all of which can serve as performance indicators and warnings of upcoming failure.

The last type of part typically labeled a frequent source of failure is a bearing.  In machines, bearings perform a number of duties – they disperse stress, limit friction, and allow rotating parts to move at the necessary speeds for great lengths of time.  However, due to the typically continuous nature of tasks which bearings must regularly do, they can degrade, wear down, or even break.

The condition and quality of bearings can be analyzed using any one of the techniques discussed above.  In addition, ultrasonic monitoring, which detects sounds too high-pitched for the human ear, can be used to quickly catch the slightest misalignment or increased friction.  The temperatures of the bearing itself and the surrounding machinery can also be measured, as in many cases parts which have degraded or are close to failure will experience unnaturally high temperatures due to enhanced friction and radiate unusually large amounts of heat.

Condition monitoring devices have been successfully developed for each of the failure modes and types discussed above. A list of some of the most frequently used devices and a brief description of their function is provided below.

Common Types of Condition Monitoring Devices

Most condition monitoring devices are hardware in nature, being pieces of equipment that either attach to or are placed nearby, on, under or in the machine itself depending on the type of data they are meant to gather.  Early condition monitoring relied on people, either within the company or hired consultants, who would evaluate the gathered data and report on the condition of the monitored machines.  Needless to say, the data analysis was not in “real time” but the improvement in plant performance for critical machines could still justify the time and expense of manual condition monitoring. 

In recent years, many condition monitoring devices have begun to be sold with accompanying software components.  Typically, this software collects the gathered data into one place and displays it in a format that is easy to read and interpret.  Condition monitoring software can also be used to plot trends and averages as well as contrasting “ideal” calculated data with “actual” observed data.  This allows for more accurate prediction of the future performance of the machine and, therefore, the ability to schedule necessary maintenance and upkeep tasks at convenient times.

Vibration Monitoring

Vibration monitoring is probably the most popular and broadly used type of condition monitoring.  This is primarily due to its versatility – most machines regularly produce or experience at least some level of vibration, making this type of device an excellent choice for companies with many different machines to monitor.

Typically, a vibration monitoring device is attached in some manner to the casing or outer part of the machine.  As rotating or otherwise moving parts within the machine perform their jobs, the casing experiences regular vibrations, which are picked up and catalogued by the monitoring hardware, typically via accelerometer technology. A number of problems can cause anomalies in the vibration pattern, including but not limited to the slowing down or stopping of moving parts, bearings rubbing against each other in an unusual manner, and any internal imbalances or misalignments.

Typically, a change in vibration will occur prior to total failure of the part, allowing repair or maintenance actions to be taken well in advance of its occurrence.  In many cases, vibration monitoring devices catch problems so far in advance that the machine can be successfully repaired without needing to be shut down for long, if at all, eliminating unnecessary, unproductive downtime entirely.  Vibration monitoring devices also save time by accurately identifying where or in what part of the machine the problem is occurring, thereby eliminating the need for often lengthy troubleshooting sessions carried out by skilled or specially trained (expensive) human operators.

Vibration analysis is the primarily recommended form of condition monitoring for any machine with rotating parts such as pumps, motors, engines, fans, blowers, and many others.  In addition, vibration monitoring devices are extremely effective in detecting misalignments of gearboxes and many types of bearings such as the pillow-box or rolling-element varieties.

Sonic and Ultrasonic Monitoring

If strange vibrations are the number one indication that something is potentially wrong with a machine, then strange sounds are certainly number two.  Worn, broken, damaged, misaligned or otherwise improperly functioning parts can make any number of odd sounds. These sounds have been described as grinding, screeching, cracking, creaking, etc.  Prior to mechanical devices to measure acoustic vibration, we have always had experienced operators in manufacturing who could “tell” a machine was ready to experience failure by just the presence (or lack) of sound.

Sonic condition monitoring devices measure and detect these sounds and then provide an alert that the machine should be inspected or maintained.  Because an odd sound is not immediately indicative of what part or section of the machine is experiencing the problem, in many cases analysis of sounds is performed along with the previously discussed analysis of vibration in order to achieve the most accurate results.

However, not all sounds produced by failing machines are capable of being detected by the human ear.  Therefore, another type of condition monitoring has recently begun to gain in popularity: ultrasonic.  Similar to the above-mentioned sound monitoring, ultrasonic devices focus on unusual sounds made by the machine that can be indicative of any number of faults – but unlike their sonic counterparts, ultrasonic devices focus on sounds too high-pitched or high-frequency for humans to hear.

The signals monitored by an ultrasonic device are expressed as data using the unit decibels per microvolt (dBuV).  Because ultrasonic monitoring focuses on both sonic and electrical elements of machinery, it is capable of detecting not only mechanical failures but electric failures as well. Ultrasonic monitoring is highly accurate and, like vibration monitoring, is able not only to identify potential failures extremely early but to narrow down and specifically pinpoint which parts are in danger of failing.  Failures capable of being detected by ultrasonic monitoring include rubbing or excess friction caused by worn-down bearings, part speeds which are too high or too slow, and unusual electrical frequencies being emitted.

Lubricating Oil Analysis and Quality Monitoring

Many types of equipment and machinery use some type of lubricating oil to keep their parts moving smoothly and prevent harsh rubbing or impacts between small, delicate elements. These oils must be carefully observed due to high risk of failure and are an excellent indicator of the equipment’s overall health.  For this reason, devices have been created which analyze the contents and composition of lubrication oils and provide extremely detailed data, usually via software.

There are currently a number of different oil analysis devices available for use, as several methods have been developed which are effective in determining the health of the oil.  A scanning electron microscope can be used to take a picture of oil samples in order to obtain close-up views of suspended debris or particles.  Spectrographic oil analysis identifies the chemical composition of the oil at various times throughout the monitoring cycle and provide alerts if unusually high amounts of certain elements or metals are present.  Wear debris detection sensors utilize similar methods but focus specifically on the detection of metal particles and the relative proportions of ferrous (iron-containing) and non-ferrous metals present in the oil.

But why do these devices all focus so intently on debris and particles? As mentioned above, the quality of a machine’s lubricating oil is frequently an indicator of its own overall quality and health. Certain types of abnormal particles found within the oil can indicate any number of potential problems.

  • Dirt, dust or grit can mean that the machine has become dirty and requires a thorough cleaning and replacement of lubricating oil. These particles can also potentially be a sign of larger problems, such as a crack or break in the machine casing allowing the entrance of harmful particles from the surrounding environment. Extremely high concentrations of dust and dirt can sometimes even indicate the failures of other machines – such as the workshop’s vacuums or cleaning system.
  • The presence of metals, including but not limited to aluminum, iron, and chrome, signifies that parts within the machine are beginning to wear down. As they rub against one another and slowly degrade, parts such as bearings, gears, cylinders and more will “shed” these particles into the lubricating oil.  By analyzing which particles are present and in what concentrations, monitoring devices can accurately pinpoint which parts of the machine are failing and what type of failure is occurring or about to occur.

It is important to regularly clean and/or replace lubricating oil to ensure that it stays healthy and continues to perform its intended function.  However, pairing these regular upkeep tasks with the installation of an oil analysis device can provide you with regular, extremely thorough knowledge of your machine’s health and warn you of any upcoming failures.

Thermographic (Temperature) Monitoring

Another common indicator that something may be wrong with your machine is temperature.  A number components, especially moving parts, tend to give off unusual amounts of heat when they degrade.  Heat can also be a warning sign that too much friction is occurring, which can be a sign of misalignments or parts rubbing against one another that should not be doing so.

Thermographic condition monitoring devices measure variations in temperature across the surface of the machine to detect areas in which unnatural amounts of heat may be being emitted. These devices are typically extremely sensitive and capable of catching small variations or abnormalities. Most devices use thermal imaging, or taking a series of “pictures” of the machine’s temperature landscape, as their central method of detection.  Infrared technology is also frequently utilized.

Because thermographic devices are also especially effective at detecting failures related to rotating parts, they are often used in combination with vibration monitoring hardware, especially on larger machines with many small, interconnected elements.

Current and Voltage Monitoring

As some of the above-mentioned devices are primarily focused on detecting mechanical failures, additional hardware has been developed which specifically focuses on electrical failures instead. These devices typically measure electrical currents running through the machine and test for unusual voltages.

Current and voltage monitoring devices can detect imbalances in the electrical supply, such as situations in which the machine is receiving too much or too little electricity to function properly, or if something has disrupted the flow of electricity or broken or disrupted any circuits. This allows for machines to be repaired before any power outrages or potentially dangerous failures, such as sparking or fire caused by an overload of electricity. Some current monitors can also detect some types of mechanical failures, such as unbalanced rotors or rotating parts moving too quickly or slowly.

Recently, a common trend among manufacturers of current monitoring devices is the utilization of model based voltage and current systems, which are believed to possess some effectiveness in detecting all three types of potential failure (mechanical, electrical, and operational). These devices compile data in order to calculate the “ideal” performance of a machine’s electrical components and/or power supply, and compare these totals to “actual” performances in order to identify areas which may be beginning to degrade.

Other Types of Condition Monitoring

The condition monitoring devices described in detail above are only the most popular and widely used of the many varieties currently available for use. Practically every parameter, property or quality of a machine can be monitored and will indicate at least something about its current health and performance. Therefore, it is impossible to list every type of condition monitoring device currently being manufactured or used. However, a few other popular examples are briefly included below.

Speed sensors measure rotating or otherwise moving parts of machines so that they can detect as soon as the machine begins to unnaturally speed up or slow down, which can quickly disrupt the flow of production.  Pressure sensors are useful for equipment that utilize combustion or flowing or pressurized fluids in their regular cycles of operation.  An imbalance in pressure, especially when it becomes too high, can result in extremely catastrophic failures up to and including explosions, these type of problems are especially important to detect prior to their occurrence.  Motor monitoring or motor current signature analysis (MCSA) devices, as the name suggests, focus primarily on the wide range of ways in which motors can fail, as the failure of a motor often leads to subsequent failure of the entire machine.

The Tactix™ System, by Proaxion®

As the list above clearly shows, there are so many ways in which a machine’s parameters can be monitored, and even more ways in which failures from the miniscule to the catastrophic can occur.  Increasingly, condition monitoring devices which focus on only a single parameter are becoming outdated, and the apparent necessity of purchasing multiple devices for a single machine expensive and complicated.  Many manufacturers have begun to develop condition monitoring equipment which observes and collects data of multiple types at once.  Foremost among these is the Tactix™ sensor, manufactured by Proaxion®, which has been praised for its 24/7 data collection, its ease of installation and ease of data use, including automated alerts to employee phones.

The Tactix sensor focuses on two major areas, both among those discussed above as the most reliable performance indicators: vibration and temperature monitoring. The system comes in two parts: a small hardware sensor which is extremely easy to install (favorably reviews as “truly plug and play”) and a cloud-based software component that organizes the collected data into an easy-to-read format. The system produces alerts as soon as changes outside reasonable limits are detected, giving employees the greatest amount of time possible to take action and rectify the issue.  Additionally the software can track trends and compare current and historical performance data.

Because it is capable of monitoring multi-axis vibration and temperature, the Tactix system has a wide range of applicability throughout manufacturing plants.  Machines and components which can be successfully monitored by Tactix include, but are not limited to, motors, gearboxes, pillow block breakers, fans, blowers, compressors, pumps, and any other rotating elements. Choosing Tactix can represent significant monetary savings as it can be used for so many different machines, compared to more limited devices or systems which focus on a single type of condition monitoring.

The system collects and stores data 24/7, eliminating the need for either planned or unplanned downtime due to maintenance. The prompt alerts save time and money as failures are detected well in advance, leaving plenty of time to develop cost- and time-efficient solutions for a wide range of failure modes. Machines monitored using the Tactix systems experience greatly extended lifespans as well as being significantly safer to operate. With its state of the art, easy to use and interpret combination of multi-axis vibration and temperature monitoring, Tactix is head and shoulders above any other devices currently available on the condition monitoring market.

But How Do I Get Started? Performing a Criticality Assessment

Once you have selected the best condition monitoring device or devices to use, however, a question still remains: to which machines or equipment should it be applied? While predictive maintenance, including condition monitoring, is the most widely useful and successful maintenance strategy, it is true that it is not the ideal strategy for every machine, and utilizing predictive tactics on certain nonessential machinery can result in unnecessary expenditures and time spent in employee training or device installation.

Several of the broader maintenance approaches, such as reliability centered maintenance (RCM) and total productive maintenance (TPM), utilize a process referred to as the criticality assessment to determine which machines would benefit from predictive maintenance and those which would not. The standard criticality assessment process divides the machines and equipment in a workshop into the three categories, sometimes collectively called the Criticality Index, enumerated below:

  1. Critical machinery.   Critical machines are extremely necessary for the day to day function of the workplace, and the production process is completely unable to continue if even one of these machines has failed or is experiencing downtime. Failures of these machines are also frequently expensive and time-inefficient to repair. These machines should always be maintained using predictive technologies, and 24/7 condition monitoring using hardware such as Tactix is highly recommended. Which machines are critical varies depending on the overall purpose of the workplace.
  2. Essential machinery.   Essential machines play an important role in the production process and the function of the workspace, but do not completely disrupt it if they should fail or experience downtime. Any machine of which a workplace contains multiple identical models (redundancy) would be considered essential. Condition monitoring and predictive maintenance are still the recommended solutions for essential machines, but, in situations where a redundancy exists, it may not be necessary to monitor every instance of the machine present in the workshop. As with critical machinery, which machines are considered essential varies according to the purpose of the workplace.
  3. Nonessential, general purpose or balance-of-plant machinery.   These are the machines which do not disrupt the production process should they fail. Equipment which improves quality of life, such as lighting systems, vacuums, and air conditioning or heating systems usually fall under this category. Predictive maintenance is not always necessary for these machines, and other strategies such as preventive maintenance or even reactive / run-to-fail maintenance are sometimes employed.

Benefits of Condition Monitoring

While each type of condition monitoring device possesses its unique advantages, as thoroughly analyzed in the earlier sections of this articles, there are a number of universal benefits which come with the implementation of a successful condition monitoring strategy. These benefits are far too numerous to list in their entirety, but frequently include:

  • Extended lifespan of machinery, as collecting data can lead to improvements which decrease the risk of future failures
  • Saving money over time due to not performing maintenance too frequently (as preventive maintenance strategies sometimes result in) or having to replace completely destroyed parts due to undetected failure (as is often the case with reactive or run-to-fail maintenance)
  • Employees who possess a greater familiarity with and understanding of the machines which they operate, due to training and the data collection, reading and interpretation involved in condition monitoring
  • Decreased health and safety risks due to failures being prevented before they occur
  • The decrease or complete elimination of both planned and unplanned downtime, as condition monitoring devices can collect the necessary data while the machines are running
  • An overall more efficient and successful production process due to the aforementioned elimination of downtime as well as a significantly lowered risk of producing deformed, damaged or otherwise unusable parts

It is clear to see from the various topics discussed in this guide that condition monitoring is a successful, efficient strategy which provides a significant number of advantages for any workplaces which choose to employ it. If you have any questions, or are interested in beginning or improving your own condition monitoring efforts via the revolutionary Tactix system, contact Proaxion today!