Helicopter HUM/FDR: Benefits and Developments. Brian D. Larder Stewart Hughes Limited Southampton, UK

Transcription

1 Helicopter HUM/FDR: Benefits and Developments Brian D. Larder Stewart Hughes Limited Southampton, UK ABSTRACT The introduction of Health and Usage Monitoring (HUM) and Flight Data Recording (FDR) systems represented a major advance in both the monitoring and management of helicopters. A considerable amount of in-service experience has now been gained and a clear picture has emerged of the benefits being provided by these systems. The paper reviews the HUM/FDR experience to date to assess the extent to which the hoped for safety and maintenance benefits have been achieved. The paper then describes some current developments which should enable further benefits to be realized. INTRODUCTION Stewart Hughes Limited (SHL), part of Smiths Industries Aerospace, produced their first helicopter HUM/FDR system with Teledyne Controls Inc. in the early 1990s. The first systems were supplied directly to civil operators in the UK and Europe and installed on helicopters operating in the hostile environment of the North Sea. Since that time the company has worked with aircraft constructors Eurocopter, Bell Helicopter Textron and Boeing, to develop other HUM system variants. SHL now has three HUMS variants in service on seven different helicopter types operating in various parts of the world. These systems have accumulated several hundred thousand hours of experience in civil and military operations. Much has been written about the potential benefits of HUM/FDR systems. The focus of this paper is to identify the benefits which have actually been achieved with the in-service systems. The paper also identifies areas where the potential benefits which have yet to fully realized and considers the reasons for this. Some current developments are then described which should both help in the achievement of those benefits which have not yet been realized, and also enable an expansion of the areas of benefit available. SAFETY BENEFITS Initially the primary driver for installing HUM systems on civil aircraft was to improve safety by reducing the rate of occurrence of accidents due to technical failures. However improving safety can generate a number of secondary benefits. For example, helping to keep down insurance costs, making helicopters more acceptable to the public as a means of transport, and providing the operator with a market place advantage. There is no doubt that the current HUM systems are providing significant safety benefits. The UK CAA have gathered and analyzed HUMS data from the North Sea helicopter operators (reference [1]). They have obtained information on 63 airworthiness related arisings, where an arising is defined as an event which has led to significant maintenance action. HUMS successfully detected approximately 70% of these arisings. The CAA classified 6 of the arisings which were successfully picked up as potentially catastrophic and hazardous, and estimated that 1 or 2 of these would most probably have led to accidents if they had not been detected in time. The CAA make the following statement in reference [2]: "It is considered that the first generation HUMS, which added comprehensive vibration monitoring to existing health monitoring techniques, has already demonstrated the ability to identify potentially hazardous and catastrophic failure modes, and has already reduced fatal accident statistics." The level of safety benefit obtained can be further assessed by considering both successful fault detections (i.e. examples of cases where a benefit has been achieved), and accidents to HUMS equipped aircraft (i.e. examples of cases of where no benefit has been achieved). Successful fault detections The most significant new health monitoring capability provided by HUMS is that of comprehensive vibration monitoring of the helicopter drive train and rotor system. Table 1 lists the SHL Presented at the American Helicopter Society 55 th Annual Forum, Montréal, Quebec, Canada, May 1999 Copyright 1999 by the American Helicopter Society Inc, All rights reserved

2 vibration health indicators and HUMS data which are referenced in this paper. Title RMS SO1, 1R, 1S, 1T SO2, 2S, 2T Indicator description The overall rms energy of the signal. The energy at once per revolution. The energy at twice per revolution. FM1A The level of low frequency modulation present in the signal. FM4A FM4B The fourth statistical moment (or kurtosis) of the signal, an indicator of localized gear tooth damage. The ratio of the 'base' energy of the signal average to the total energy. BBand Broad band vibration analysis, trending vibration levels in multiple frequency bands. Sig Av Trk & lag Diagnosis aided by an examination of the raw vibration signature. Rotor blade track and lag data. Table 1: Referenced vibration health indicators SHL maintains a database of HUMS fault case histories, this contains a large number of transmission and rotor system faults which were either detected by the HUMS vibration diagnostic function, or for which significant data trends were identified. Table 2 lists a selection of the different types of fault for which significant HUMS indicator trends have been identified to illustrate the variety of faults which have been detected. The table also identifies the HUMS indicators which provided information on each of the fault types. No Fault category Gears 1 Bull pinion and wheel gear: Abnormal gear mesh HUMS Indicators FM1A, FM4B 2 Bevel gear: Tooth defect FM4A 3 Bevel pinion and wheel: Tooth spalling/ pitting 4 Combiner gear: Loose and worn retainer ring Shafts and couplings 1 Cross shaft: Coupling bolt fatigue failure 2 MGB input shaft: High vibration RMS SO1, SO2, FM1A 1S 1S, 2S No Fault category 3 MGB through shaft: Excessive wear on splines 4 Tail drive shaft: Out of balance/ misalignment 5 Tail drive shaft: Incorrect shaft section installed 6 Oil cooler fan drive shaft: Failure 7 Oil cooler fan: Out of balance from sand deposits Bearings 1 Fan drive thrust bearing: Spalling and cage damage 2 Main bevel shaft support bearing: Outer race crack HUMS Indicators SO1, SO2 1S 1S SO1 SO1 SO2 FM1A, FM4B, Sig Av FM4A, Sig Av BBand 3 Main bevel shaft support bearing: Loose 4 Oil cooler fan bearing: Damaged 5 Tail drive bearing: Wear/ BBand raceway damage 6 Tail drive bearing: Requires BBand lubrication 7 Tail drive bearing: Detached BBand seal 8 Sync shaft bearing: Cracked 2S housing Rotors 1 Rotor lead/lag dampers: Defective 2 Tail rotor flapping hinge bearing: Damaged 3 Tail rotor pitch link bearings: 1T, 2T Damaged 4 Tail rotor pitch link bearing: 1T Seized Miscellaneous 1 Ring gear: Loose split line bolts RMS 2 Free wheel unit: Wear on cams RMS, SO1, FM1A 3 Rotor brake: Loose on drive splines 4 Rotor brake module: Out of balance 5 Accessory module: Bearing and gear tooth damage 6 Blower drive: Broken rivet nuts and flange damage Table 2: SHL HUMS fault cases Trk & Lag, 1R 1T FM4B SO1 FM4B FM4A

3 Reference [3] presented summaries of ten example HUMS fault case histories, new fault cases have been added to SHL's database since that paper was written. To illustrate the range of experience being gained, examples of three of the more unusual new fault cases are given below. Example 1, Abnormal gear mesh: The FM1A indicator for one of the pinions driving the main rotor gearbox bull gear showed a slow upward trend over a period of approximately 800 hours, and eventually crossed the monitoring threshold (Figure 1). On the bull gear the FM1A indicator showed an upward trend over this period and FM4B also steadily increased. A boroscope inspection revealed that the pinion and bull gear had an abnormal tooth wear pattern. In one location the tooth crown of the pinion gear had bottomed out in the root radius of the bull gear, creating a marked wear step. The gearbox was rejected and returned to the constructor, who subsequently confirmed nonconforming dimensions for gearshaft runout and the profiles and pitch of several teeth. Figure 2: Bull pinion FM4A trend Example 3, Tail drive bearing vibration: This is not strictly a fault case, but is interesting nevertheless. Figure 3 shows the high frequency vibration generated by a tail drive shaft bearing. This has a 'saw tooth' trend, with progressive increases in vibration followed by step decreases. The reason for the trend was not initially understood, however questions to the operator revealed that the step decreases in vibration correlated with the periodic regreasing of the tail drive shaft bearings. The rising vibration trend could be used to indicate when bearing re-greasing is required Figure 1: Bull pinion FM1A trend Example 2, Broken rivet nuts and flange damage on blower drive: The FM4A indicator for one of the pinions driving the main gearbox bull gear showed a sudden increase (Figure 2). The decision was made to perform a boroscope inspection. When the blower drive cover was lifted (the blower drive is not used on this aircraft), severe damage was found on the cover inner surfaces and cover adapter. In addition four rivet nuts had broken off the blower drive flange and the drive flange was damaged. The blower drive assembly was removed and the gears inspected, no gear damage was observed. The FM4A trend subsequently returned to normal levels. The operator attributed the increased FM4A trend to the impact of rivet nut fragments against the inner surfaces of the blower drive cover Figure 3: Tail drive bearing: High frequency vibration trend The number and variety of fault cases referenced above illustrates that the comprehensive HUMS vibration monitoring function provides an effective drive train and rotor system fault detection capability. Accidents to HUMS equipped aircraft Since HUM/FDR systems were introduced to the North Sea helicopter fleet approximately 8 years ago, there has been one (non UK) fatal accident and one (UK) serious incident due to a technical failure on a HUMS equipped aircraft (neither were fitted with SHL HUM systems). In both these cases the HUMS could have prevented the accident/incident, but failed to do so - for a different reason in each case.

4 The fatal accident occurred on Super Puma LN-OPG in September This was the result of a failure of the right hand main gearbox input assembly which resulted in an overspeed and disintegration of the No. 2 engine. The fault on the right hand main gearbox input assembly was potentially detectable by vibration monitoring. There were two key HUMS sensors monitoring the main gearbox input and engine which could have provided warning of the failure. Unfortunately one of the sensors was unserviceable in the period prior to the failure, whilst the other was used only for data recording and not for monitoring (i.e. with thresholds and automatic warnings). The serious incident occurred on Super Puma G- PUMH in September 1995 and was the subject of a UK Air Accidents Investigation Branch (AAIB) investigation (reference [4]). A tail rotor blade flapping hinge retainer had fractured, almost resulting in the loss of a blade. The helicopter's HUMS began to identify a developing vibration problem some 50 hours previously, culminating in an associated exceedance alert some 5 hours before the incident. However maintenance activity to identify the cause of the tail rotor vibration problem failed to detect the cracked retainer. Therefore the crack propagated undetected until a near catastrophic failure occurred, resulting in the onset of severe airframe vibration in flight. Both the accident and incident highlight some important issues. Neither can be attributable to a failure of the HUMS and indeed they illustrate the potential contribution to safety it can make. They do however highlight the weaknesses in the HUMS implementation and regulatory environment which have to date prevented this potential from being fully realized. In the 'end to end' process weaknesses exist at all of the stages from the targeting of critical failures to reliably monitoring those failures and producing meaningful information which enables positive intervention in the maintenance of the aircraft. Excluding the Super Puma Mk 2 and EH101 (which are certified to BCAR 29), there are currently no regulatory requirements for HUMS which apply to the aircraft now fitted with systems. Therefore there is no agreement as to the objectives of HUMS. Whilst the regulators see HUMS as a safety tool, the constructors will not accept this as the aircraft have been certified as airworthy without HUMS. Without agreed objectives, there is no agreed definition of the requirements for, and criticality of, different HUMS functions. If the criticality of the HUMS functions has not been defined, system serviceability requirements, operational limitations and compensating actions cannot be defined as they are for all other aircraft systems (MEL etc.). Finally, there cannot be agreed requirements for the management of information generated by HUMS and the aircraft maintenance procedures associated with this. All the above issues have until now been left to the helicopter operators alone to address. What is clear is that if the potential safety contribution of HUMS is to be fully realized then the regulators, aircraft and system manufacturers have to play an equally active role. All of the concerned parties have a contribution to make and it is important that these contributions are facilitated. A leading role being taken by the regulators is a natural and essential development. Summary of safety benefits HUMS has provided substantial safety benefits, and there has been a notable reduction in North Sea helicopter fatal accident statistics since the systems were introduced. However these benefits are currently being limited by weaknesses in the environment in which systems are operating. MAINTENANCE BENEFITS Achieved benefits There is no doubt that HUMS is providing worthwhile maintenance benefits, even though there is little quantitative data on achieved cost savings. Civil operators do not have the time or resources to gather HUMS related cost data, and cannot afford the luxury of being able to carry out controlled evaluations. Combining in-house experience (reference [5]) with that reported by helicopter operators, (e.g. in reference [6]), achieved maintenance benefits can be summarized as follows: Simplified rotor track and balance procedures and reduced test flying for rotor track and balance. Reduced airframe vibration levels, resulting in fewer avionic and other faults, and hence less maintenance and downtime. More pro-active maintenance, as a result of being able to track developing HUMS indicator trends and take appropriate actions in a timely manner.

5 Better targeted maintenance as result of more accurate exceedance information and improved troubleshooting. As a result of the above items, reduced disruption to operating schedules due to unscheduled maintenance and aborted flights. Benefits still to be achieved The primary area in which benefits have still to be realized is that of 'maintenance credits'. Reference [7] defines the award of a credit as "to give approval to a HUMS application that adds to, replaces, or intervenes in industry accepted maintenance practices or flight operations." A credit can be either a 'health credit', based on more extensive component health information provided by the HUMS, or a 'usage credit', based on improved component usage information. A Maintenance Credits Working Group (MCWG) issued a report in November 1992 (reference [8]) which stated: "Maintenance credit potential exists for all major dynamic components of rotorcraft, for mechanical and hydraulic components of flying controls, and for critical parts of the airframe and undercarriage. Considerable scope exists for relaxing maintenance tasks at all phases of component installation, service operation, and reconditioning, where supported by effective HUM." Despite the above statement, to date few maintenance credits have actually been introduced. Again, this is not necessarily the fault of the HUM system. Rather it is due to the fact that HUMS is a major development which takes time, effort and the involvement and cooperation of multiple organizations, to fully integrate with established maintenance and life control practices. Aircraft constructors must understand the capabilities and limitations of the HUMS functions applied and identify credits based on this. Regulatory bodies must certify the HUMS and validate any credits. Summary of maintenance benefits Worthwhile maintenance benefits have been obtained from HUMS, however the potential for maintenance credits has still to be realized. This is mainly due to the fact that HUMS needs the development of supporting processes for the identification and approval of maintenance credits. HUM/FDR DEVELOPMENTS achievable from HUM/FDR systems. These include initiatives being taken by the regulatory bodies, activities of the helicopter constructors, and developments in system functionality. Activities of regulatory bodies A regulatory framework is being put in place to enable HUMS to realize its full benefit potential. The UK CAA has been progressing a pro-active policy for the installation of HUMS for new (BCAR 29) certifications. In line with this policy, the CAA has adopted new requirements for the certification of large helicopters which include a rotor and transmission design safety assessment, in which health monitoring is an acceptable compensating provision. The first two helicopters to be certified to the new requirements are the Super Puma Mk 2 and the EH101. Now the UK CAA intend to issue an Airworthiness Directive (AD) which makes the installation and use of health monitoring systems mandatory for all UK registered commercial air transport rotorcraft certified to carry more than 9 passengers (reference [2]). The decision to mandate the installation and use of health monitoring systems was taken in response to the AAIB's recommendations relating to incidents and accidents such as G-PUMH. The draft AD specifies the following requirements: 1. Install a health monitoring system approved for the type (where not already installed). 2. Implement adequate procedures covering all aspects of data collection, analysis and determination of serviceability. Where an approved system is fitted, the operator will need demonstrate compliance with requirement 2 by submitting information which includes the following subjects: Duties and responsibilities of health monitoring system personnel General vibration monitoring process procedures Health monitoring system facilities Procedures for changing system equipment, software or maintenance practices Threshold setting and adjustment Limitations Training This section of the paper reviews some current developments aimed at maximizing the benefits

6 The CAA is consulting with aircraft constructors and operators prior to finalizing the wording of the AD. In parallel with the above activities, a harmonized committee named the RHUMSAG (Rotorcraft Health Usage Monitoring System Advisory Group), with members from the FAA and JAA and other bodies, is developing Advisory Circular (AC) material for the airworthiness approval of HUMS (reference [7]). The AC establishes an acceptable means for certifying a rotorcraft HUMS. It defines certification requirements in terms of equipment installation/ qualification, credit validation, and instructions for continued airworthiness. The material has been developed in response to an increasing number of certification applications to install HUMS and use its data to provide, amongst other things, maintenance credits. Activities of helicopter constructors All the major helicopter constructors now have active HUMS programmes and are working to extend the benefits of HUMS, as the following examples show. Sikorsky and Bell Helicopter Textron are involved in the development of HUMS for new aircraft types such as the S92 and the BA609. SHL are working with Eurocopter on the integration of HUMS outputs into the aircraft maintenance manual. Eurocopter have identified potential health credits from HUMS, including the elimination of some inspections, extension of inspection intervals and extension of gearbox overhaul periods. Bell have developed a system known as BUCS (Bell Usage Credit System) which has the capability to provide usage credits in the form of extended component lives. Development of HUM/FDR functionality In partnership with a number of other organizations, SHL are involved in a range of HUM/FDR developments, for example: New systems, with functionality tailored to the requirements of aircraft type and size. New monitoring functions, for example the use of electrostatic sensors for engine gas path and oil system monitoring New active functions such as collective limit cueing to eliminate torque exceedances and allow helicopter pilots to fly "eyes-out" of the cockpit at times of high workload. This paper briefly introduces one current development in the UK which represents an expansion in functionality that has the potential to offer benefits in a completely new area. Traditionally the FDR part of HUM/FDR systems has been a reactive tool for accident and incident investigation. However, with the active encouragement of the UK CAA, airlines such as British Airways have, for many years, been using flight data in a pro-active flight operations monitoring programme. The objective of this programme is to detect adverse trends in flight operations to enable action to be taken to prevent incidents and accidents from occurring. More recently, a similar effort has begun in the US, in what is known as the Flight Operations Quality Assurance, or FOQA, programme. The airline industry has demonstrated that flight operations monitoring programmes can improve safety, whilst providing a number of other operational and economic benefits. These programmes monitor flight operations by routinely analyzing aircraft flight data to detect 'events', which are deviations from normal, expected, or flight manual practice. They provide continuous operational quality control with timely feedback on sub-standard practices, and produce valuable information for the evaluation and improvement of procedures. Now that all large UK public transport helicopters are equipped with HUM/FDR systems, the CAA has initiated a research effort which is implementing a flight operations monitoring programme on helicopters. This has been called a Helicopter Operations Monitoring Programme (HOMP). SHL has been at the centre of this research effort, the company has carried out a feasibility study on the implementation of flight operations monitoring on helicopters and is now prime contractor for the execution of a trial programme. Helicopter flight operations monitoring feasibility study The feasibility study investigated the potential safety and other benefits of a HOMP based on the analysis of an existing set of flight data from one helicopter, gathered over a period of one year. The analysis showed that it was possible to detect a range of helicopter operational 'events'. The flight data trace for one such example, which is that of the helicopter performing a relatively severe 'quickstop' manoeuvre, is shown in Figure 4.

7 Events: High ROD, High NR power off, High pitch attitude Figure 4: Flight data trace for example HOMP 'event' The study also demonstrated the ability to generate outputs which characterized the operation and showed event trends. Example outputs are presented in Figure 5. Flight envelope characterisation Event trend analysis Distribution of maximum pitch rate values % Occurrences of NMLA events Frequency % 80.00% 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% No of occurrences Dec- Feb Mar- May Jun- Aug Sep- Nov Events Sectors % Period of year Upper limit of Bin Event 21D: High roll rate above 500ft AGL In-flight handling: Ranking of events by number of occurrences (sectors) 14 EVENT22A High Rate of Descent below 500 ft AGL EVENT75A Max Continuous Torque (2 Engines) EVENT7A Low Airspeed above 500 ft AGL 12 EVENT80 Autopilot Engaged below 70 knots EVENT20E2 High pitch rate below 500ft AGL Frequency EVENT22C High Rate of Descent above 500 ft AGL EVENT20B Low Pitch Attitude below 20 ft AGL EVENT20A High Pitch Attitude below 20 ft AGL EVENT20D Low Pitch Attitude above 20 ft AGL EVENT21 Roll Attitude EVENT20E High Pitch Rate above 500 ft AGL EVENT21D2 High Roll Rate below 500 ft AGL EVENT23B Airborne Normal Acceleration (max) EVENT21D High Roll Rate above 500 ft AGL EVENT20C High Pitch Attitude above 20 ft AGL 2 EVENT74A Excessive Lateral Cyclic Control EVENT81B Gear not locked down below 500ft (ldg) Upper limit of Bin 21 EVENT23B2 Airborne Normal Acceleration (min) EVENT81A Gear not locked down below 100ft (t/o) EVENT73B High Rotor Speed - Power Off EVENT84 VNO Figure 5: Example 'trend' outputs

8 As a result of the success of the feasibility study, the work has now been taken forward into an operational trial. Trial helicopter flight operations monitoring programme With the sponsorship of the CAA and Shell Aircraft Limited, a HOMP trial has been set up, involving 5 Super Puma helicopters operated by Bristow Helicopters Limited. SHL are the prime contractor for the trial. The helicopters will be fitted with PCMCIA Card Quick Access Recorders (supplied by Marconi Avionics) for the daily downloading of the FDR data. The data will be processed in a flight data replay and analysis system which is being supplied by British Airways. This will be based on the Flight Data modules in BASIS (the British Airways Safety Information System). Once equipment development, installation and commissioning is complete, Bristow Helicopters will run the trial programme for a minimum period of one year. CONCLUSIONS HUM/FDR systems have generated significant safety benefits. They have detected a number of potentially catastrophic or hazardous failures and have almost certainly prevented accidents. Systems are also providing worthwhile maintenance benefits. To date, the achievable benefits have been limited by the need for aircraft constructors to be fully involved in the system implementation and for the regulatory bodies to develop an appropriate regulatory environment. [2] UK CAA Draft CAP "Acceptable Means of Compliance, Helicopter Health Monitoring". Draft Issue A, January [3] Larder B. D. An analysis of HUMS Vibration Diagnostic Capabilities. American Helicopter Society 53 rd Annual Forum, May [4] AAIB Aircraft Incident Report 2/98 "Report on the Incident to Aerospatiale AS332L Super Puma, G-PUMH over North Sea on 27 September 1995". [5] Larder B.D. "Update on HUMS Transmission Vibration Monitoring Experience". Institution of Mechanical Engineers Seminar S553, November [6] Dobson I. "Transmission Health Monitoring Experience from Rotorcraft Operated by Bristow Helicopters". Institution of Mechanical Engineers Seminar S553, November [7] JAR-29 NPA 29-18, containing Draft AC Material "Airworthiness Approval of Health Usage Monitoring Systems". October [8] Helicopter Health Monitoring Advisory Group Working Group Reports: "Maintenance Credit, Usage Monitoring, HUMS Implementation". CAA, London, February Regulatory bodies, aircraft constructors and HUMS suppliers such as SHL are involved in a number of developments which should enable all of the hoped for benefits to be realized. In addition, new developments such as the pro-active use of FDR data should enable some entirely new benefits to be generated. ACKNOWLEDGEMENTS The author wishes to thank the UK CAA, Eurocopter, Helikopter Service A/S, and Schreiner Northsea Helicopters, for providing information for this paper. REFERENCES [1] McColl J. "Overview of transmissions HUM performance in UK North Sea Helicopter Operations". Institution of Mechanical Engineers Seminar S553, November 1997.