Reliability Management for Helicopters: How to make the transition from preventive to predictive maintenance
Helicopters traditionally require far more maintenance per flying hour than fixed wing aircraft because of greater structural fatigue. In addition, helicopter maintenance programs are often based on MSG-2 principles with high levels of preventive component replacements. In recent years, regulatory authorities are releasing guidance material on the transition from an MSG-2 to MSG-3 based approach for twin engine helicopters. This requires the core principles of reliability management to be performed as part of the approved maintenance program. In this EXSYN insight we dive into how Helicopter Operators can make this transition and how to harness the data required for effective reliability management.
Background
The Bell 429 was the first helicopter to be certified through the MSG-3 process in an attempt to optimize costs and maintenance time. Helicopters traditionally require far more maintenance per flying hour than fixed wing aircraft as a result of greater structural fatigue owing to the cyclic nature of their operations.
Regulatory authorities are releasing guidance material on the transition from an MSG-2 to MSG-3 based approach for twin engine helicopters. Having seen the impact that this had on fixed wing aircraft when the transition occurred, it is safe to assume that this process will be far more cumbersome for rotor wing machines.
Subsequent to the Bell 429, Eurocopter and AgustaWestland have designed the EC175 and AW189 to be compatible with the MSG-3 process. OEMs are also placing more sensors on the helicopters to collect information to support this paradigm shift of moving from time-based to condition based maintenance. Bell became the first OEM to place RFID tags on all the parts of the 525 to simplify the access to technical data pertaining to parts. A combination of data from the Vehicle Health Monitoring Systems and the HUMS span across the mechanical systems (former) and the electrical and avionics systems (latter) to paint a picture of the health of the helicopter.
But even before different helicopter models are subject to this transition, operators currently flying their helicopters on an MSG-2 hard time design can harness their component data to experience savings even if they aren’t necessarily to the tune of the 15-30% that a transition to MSG-3 asserts.
The fundamental idea behind the hard time approach is that the reliability of a component decreases with increasing operating age. It is a preventive process. But a closer look at your data can assist in reducing unplanned maintenance downtime and improvements in availability.
Let us take an example of a factor very prevalent in helicopter operations and maintenance: vibrations.
Vibration control is not only essential for passenger comfort (with anti-vibration systems going into all helicopters) but also to reduce vibration induced maintenance requirements. The impact on vibrations on, say, a rotor hub has driven OEMs to use thermoplastics as an alternative to metals as damaged induced cracks propagate slower than in conventional metallic ones. The removal criteria for rotor head parts are determined by the depth of the cracks. An approach to understanding a combination of historic maintenance data coupled with sensor data is a leap towards condition based maintenance.
Let us take a look at how this can be templated:
Vibration Data – Spectrum, Monitor, Signature
Log Sheets & Inspection Data
Fault History Data & Exceedances
Load Penalties applied (as part of the broader Technical Records data)
Scheduled & Unscheduled maintenance history
Scheduled & Unscheduled removals
Analysis of such data would present broad trends across components qualitatively:
Quantitatively assessing this data to support failure modeling requires mapping of historical faults, removals (scheduled and unscheduled) and nature of maintenance carried out.
Failure modelling with historical and baseline engineering data acts as a premise to plot broad trends pointing towards predictive maintenance for those components that allow it. Every helicopter is operated and maintained differently, and therefore the historical data benchmarks which failure model best represents individual components. This is further appended with HUMS and VHMS data to fine tune to the findings.
Where Do You Start?
1. Create & Assimilate the Historical Database
The first step towards initiating this process is to dig into your maintenance and technical records archives – a combination of information residing on:
Paper
Scanned Documents
Electronic formats (PDF, Excel etc.,)
Maintenance & Engineering / MRO Systems (Swiss Aviation Software AMOS, Rusada Envision / RAL, Ramco Aviation, AMS, TRAX, IFS / MXi etc.,)
An organic database consisting (almost) back-to-birth traceability, wherever available, will be the starting point of this exercise.
2. Audit and Assimilate
The first point will present datasets that are duplicate / redundant in nature with conflicts between what is stored on paper and what is stored electronically. This includes verification of embodiment of applicable Service Bulletins or MODs against the whole regulatory database.
Certain helicopter types present challenges in terms of calculation of load penalties and other factors intrinsic and unique to specific models. Some of the Aviation Maintenance & Engineering / MRO software do not fully cater to these needs.
Examples include:
Penalty loads on AW139 when operating under wind speeds more than 27 knots, or a weight above 6400 kg
Retirement Index Number (RIN) & Accumulated Total Cycles (ATC) against Bell models
… etc.,
Ensuring accuracy of all parameters through audits and semi-auto verification tools using technologies is a critical phase.
3. Link Financial Data
The financial / cost data against maintenance carried out (matching invoices from your financial systems / source data), procurement against scheduled & unscheduled removals (this will be a large, varied set of data depending on the part type, procurement type – AOG loan, purchase, exchange) plays a role in factoring economic ageing of components in addition to technical ageing due to wear and tear.
IT support to interface with financial systems (e.g. SAP, Oracle, Navision etc.,) and trace it back to the source (maintenance type, procurement type, reason – AOG / others) is essential to factor in these variables.
What Next?
Failure analysis leading towards a predictive maintenance ecosystem requires the database to feed into mathematical models that categorize component behavior based on probability distribution functions to derive the behavioral “best-fit”.
The three stages of incremental scaling include feeding in:
Historic Data (as discussed in the previous paragraphs) & projected utilization forecasts
METAR data (impact of weather and operational environments on ageing)
VHMS & HUMS data (if available)
Adding financial / cost data allows decision economic decision making between deciding if a part is best preventively replaced or rather allowed to fail.
How EXSYN can help?
EXSYN’s experience in dealing with data audits, interfaces, migration and fleet specific knowledge allows us to tailor a custom approach towards data consolidation.
In addition, EXSYN’s experience with a varied array of Maintenance & Engineering / MRO systems (including the like of Swiss Aviation AMOS, TRAX, IFS MXi, OASES Commsoft, AMASIS, Rusada Envision, Ramco, Pentagon 2000 and other legacy systems etc.,) allows us to interface, audit and tap into the right data sources required for such an exercise.
This data set is initially run through the models of our AVILYTICS suite of solutions to plot baseline trends of component failure against each identified component & ATA Chapter. The learning algorithms, through iteration, present the results across different models to identify the “best-fit” that pertains to each serial number.
Coupled with financial / cost data, the system plots trends to support decision making around preventive removals or allow to fail.
The entire interface between the required systems is setup to avoid repetition in the future, unless there is a change in systems or an upgrade that fundamentally requires a remap.
The initial analysis is done with the maintenance history data alone. Subsequent to understanding the nuances of each helicopter and component, the additional data from VHMS & HUMS can be loaded for fine tuning the results.
Do you have any questions?
Feel free to contact us or give us a call on +31 20 8200 7600. We would be happy to help you further.
What is AVILYTICS?
AVILYTICS is a fully out-of-the-box aircraft reliability management solution that focuses on providing insights in technical reliability, upcoming potential technical failures as well as organizational efficiency analytics. It combines the traditional scope of aircraft and fleet reliability management with advanced techniques from predictive analytics to also build AOG risk profiles of aircraft, identify aircraft based reoccurring defects and measure organization performance. A full holistic approach to using data in order to increase aircraft availability and fleet performance.