Predictive Maintenance

Predictive facility maintenance is a strategy that is based on the early detection of system failures, so that it is possible to change the affected component during a scheduled shutdown, before the failure occurs.

Among the existing technologies for prematurely detecting faults, the one based on acoustic emission is perhaps the one with the most potential both in a variety of applications and with detection capability, but it is still in a very initial state of its development due to the complexity of the phenomenon to be detected.

Physical principle

Any modification in the state of a material such as the appearance and growth of cracks, local plastic deformation, corrosion or phase changes, etc. is accompanied by a release of energy in the form of pulses of elastic waves that propagate within the material. These pulses contain information from the source that generated them, and the EA technique is based on the detection and analysis of these. The advantages of this strategy over more established ones are:

- Ability to detect multiple phenomena in early stages

- Ability to detect invisible faults

- Ability to detect under normal operating conditions

- Ability to determine the origin of the failure

The release of vibratory energy is different for each phenomenon and for each material, with the identification of the parameters of the vibrating pulses associated with each type of fault and material being the main difficulty and challenge of this strategy.

Signals from an AE system.

Applications

Industrial applications of EA are still limited:

- Inspection of pressure vessels

- Detection of fluid leaks in valves and pipes

- Corrosion detection

- Bridge monitoring (under development)

But research on this topic shows great potential for application in other fields:

• Degradation of materials: increasing defects, crack propagation, plastic deformations, rupture of inclusions or precipitates, surface degradation [3], both in structures and mechanisms. This gives it great potential for application in the aerospace industry, in the processing industry, in infrastructure maintenance or in wind turbines.

• Reversible processes: phase crystallographic transformations, solidifications, thermoelastic effects, friction between surfaces, etc.

• Manufacturing processes: defects in welding, forging processes, milling, turning, etc.

• In fluids: detection of suspended particles, evolution of the gas, boiling point.

• Ability to predict the time between the detection of the onset of the fault and the macroscopic fault.

• Understanding the breaking mechanisms in non-homogeneous materials, such as steels or composite materials.

Experience

The LEAM has worked successfully with the characterization of the failure mechanisms in tool steels in order to design a predictive maintenance system in forming tools within the MOUSICOSYS project. EA has been applied in monotonic flexion, flexion fatigue, contact fatigue, and nanoindentation assays.

The ability to detect damage to biological material has been demonstrated, opening the door to the detection of injuries and programming of high-performance training or in concrete-based composite matrices, allowing to identify the mechanisms of occurrence of fail.

And finally, within the framework of the LOOMING FACTORY project, a low-cost system for the acquisition and analysis of EA-based faults and application to gear trains is being developed, which could be autonomous taking advantage of parallel development. in energy harvesting based on the transformation of vibratory energy.

Examples of internal fractures detected by AE in steels. Note the size of the fractures detected.

AE testing on organic tissues.

Detection of faults in concrete-based composite material. A vibrating pulse can be seen on the screen.

Monotonous bending test in steel test tubes for tools.