Every industry installation is set with distinct types of machinery according to its purpose. Each installation will build its own history by dealing with and maintaining assets.
One may say that machines produce signs and can actually warn, if sensored, that something is wrong. When machinery fails it generates costs with spare parts, with labor man-hours, and eventually, production downtime. Machines may fail for diverse reasons that, depending on the situation, may be harder to detect.
A machine ‘loss of usefulness’ is originated from three main sources:
- Surface degradation;
Surface degradation is by far the main cause of machine unavailability. It is also a reason for increased vibration.
POTENTIAL CAUSES OF MACHINE FAILURES
Vibration increase and excessive component heating are signs that something is going in the wrong direction. In the case of non-typical vibrations, the main causes are: defective bearings and gears, electrical and mechanical failures in motors, problems caused by misalignments, unbalances or unstable bases, bent shafts, pulley or belt failures, mechanical gaps, aerodynamic or hydraulic problems, etc.
Aside of non-typical vibration and heating, other failure indication sources are water and oil leaks, corroded pipes, smells and abnormal sounds.
Lack of or inadequate lubrication and not attending equipment safety stops when recommended can be the cause of failures. Incorrect operation of the machinery may result in malfunctioning of its internal parts and subsequent failure.
In addition, the lack of or improper maintenance can cause accidents or even machine breakdown, resulting in injury risks to the operator.
HOW TO IDENTIFY MALFUNCTIONS IN YOUR INDUSTRY
Malfunction or failure identification starts with a well-trained team. In addition to selecting qualified candidates, team training is a never-ending task. In order to obtain the best from a machine operator and maintenance team, a continuous learning process shall be established. There is no set formula. Quality work is about knowledge retention and sharing among team members of the organization. It shall comprise on the job learning added to formal education and supervision.
Documented processes and tasks, be it machine operation or maintenance, should be kept up to date and with easy access to those who need them.
Is it possible to think while putting out the daily fires?
Failure analysis capability should be encouraged for maintenance staff. Setting priority to analyzing the most relevant problems and systematically applying this procedure will generate benefits for workers with diverse functions and experience.
Root cause analysis (RCA) is a problem-solving method that consists of investigating an incident or failure, however, since it is not possible to address every situation, a trigger for a RCA is necessary. A predefined parameter to trigger the root cause analysis technique. This trigger will vary over time, and from industry to industry. The root cause analysis method can be used when:
- An unexpected event happened causing worker’s injury or death;
- A situation of unacceptable risk or repeated failure occurs;
- Machinery stops for a set time and cost. Example: over two-hour stops.
It is not enough to identify the root cause of a problem. It is necessary to plan, execute and follow on in order to verify the effectiveness of the intervention.
Just as malfunctions and failures have varying causes, its detection may be possible by routine inspections or require the application of specific techniques. Each industry sets its own approach to machinery maintenance: there are situations where running to failure could be appropriate; or adopt preventive maintenance, scheduled over a time period commonly as indicated by the OEM or by maintenance best practices will be most appropriate.
Greater industry competition is demanding advanced steps in regard to asset maintenance. That is so to reduce costs, increase machine availability and improve reliability to ensure product quality and timely shipping. Predictive maintenance or maintenance based on asset reliability steps in right here. Some of the predictive techniques are: vibration analysis; oil analysis; ferrography; thermography; ultrasound; temperature.
There is a reason to adopt the best maintenance strategy for each type of equipment. And, it is directly related to the use of resources such as spare parts and labor allocation as well as considerations about the risks that critical equipment failures represent to the organization. Overall impacts of production downtime, risk of accidents and environmental impacts shall be considered prior to adopt a given asset maintenance strategy.
PARADIGM BREAK WITH SPECTRAL AND TEMPORAL ANALYSIS
When it comes to vibration, each component of an operating machine vibrates differently, generating its own frequency spectrum. Thus, the analysis of the spectrum is one of the predictive maintenance techniques used to determine the condition of the machinery under evaluation. Through it one can identify abnormalities in the vibration of the components. In addition to detect early or advanced stage failures, those abnormalities may point out: a) the location of the affected components; b)indicate the cause or source of the problem, and; c) suggest trends that may indicate when the identified problem will become critical.
In general, to perform spectral analysis, which is obtained through FFT (Fast Fourier Transform), it is necessary to hire specialists. It is a known fact that these specialists are expensive and their time is limited. Therefore these collections are often done by sampling, which does not necessarily generate a realistic picture of the machinery condition.
In addition to spectral analysis, it is also useful to perform continuous temporal monitoring of vibration (and temperature) by devices with acceleration (and temperature) sensors. This monitoring generates time series that indicate trends of the component condition over the hours or days, suggesting to the maintenance team an in depth study by doing a spectral analysis and intervention, if required.
Following the logic of the maintenance strategy adopted for each machine based on its criticality, for its simplicity and low cost, industry 4.0 technologies allow the monitoring of several points in the same equipment in a continuous way. It’s done by recording time series and performing spectral analysis whenever demanded at the monitored point.
This is DynaPredict’s value proposition: Bluetooth data logger with acceleration and temperature sensors, wireless device and three years (use dependant) battery life. It records temperature, acceleration and RMS velocity and also performs a spectral analysis.