Extending HUMS to rotor systems

London
Source:
This story is sourced from Flight International
Subscribe today »

There is little doubt that health and usage monitoring systems (HUMS) have brought huge benefits to helicopter safety - the single greatest advance ever in the eyes of many in the field.

Yet even in the HUMS era there have still been catastrophic mechanical failures and plenty of close calls in which sometimes only highly skilled airmanship or sheer good fortune saved the aircraft and lives.

A key reason for those accidents and incidents was that the failures were manifested not in the drive-train, where HUMS has become so effective, but in the main or tail rotors themselves, which currently lie largely beyond the reach of HUMS sensors.

cpl andrew saunders caf 
 © Cpl Andrew Saunders CAF

Now the UK Civil Aviation Authority is revisiting the question of whether health monitoring can be extended to cast its electronic eye over the rotor systems and, just as significantly, whether the investment involved is worth the likely benefits.

The agency has published the results of research performed by GE Aviation, now a key HUMS player through its acquisition of Smiths Aerospace, and funded by the CAA's Safety Regulation Group, trade association Oil & Gas UK, and Shell Aircraft in CAA Paper 2008/05 called HUMS Extension to Rotor Health Monitoring.

FEWER ROTOR FAILURES

The project found that it is doubtful whether some faults could be detected at a useful point even with new technologies, and that the declining number of rotor failures significantly reduces the potential benefit.

But it also concludes that there are potential gains, and that further research is worthwhile. Some of that research is now under way and the CAA, which is particularly intrigued by the possibilities of applying HUMS techniques to tail rotors, is considering carrying out more work.

Research project manager David Howson says: "For tail rotors, it is considered that there is potential in applying the anomaly detection techniques successfully applied to transmission HUMS to provide reliable warnings in a timely manner. Research into this is already under way. For main rotors, the situation is not so clear.

"For tail rotors, there is evidence that good use could be made of the data already captured. For main rotors it may be that different parameters are required, and/or that current parameters need to be measured under different conditions to present, and/or that current parameters need to be measured at a different location, for example on the rotor rather than on the fuselage as at present. Further research would be required to establish the most fruitful way forward."

Some data that is already collected by HUMS, particularly for rotor-tracking and balance, might itself be useful and is not currently analysed. Howson suggests that there is potential for further investigation and that, again, the best hope lies with tail rotors.

His view about the emphasis on tail rotors versus main rotors is heavily driven by the accident and incident record (see panel). Not only does the data have to be captured, but in a timeframe and with appropriate operational procedures that would make it feasible for maintainers to act on it.

DETECTABLE FAULTS

In the case of some accidents, even though the underlying faults may have been detectable a short period in advance, it is questionable whether the data could have been acted upon.

Howson says: "There is cause for cautious optimism of success in the case of tail rotors, but it is not possible to provide a sensible answer for main rotors at this time. It seems likely that success may be feasible for some types of main rotor defect, but not for others."

A major specific difficulty with monitoring rotors is the technical issue of devising sensors and databuses that can collect and transfer data from the rotating modules to the fixed fuselage.

But new technologies are giving cause for hope. Howson explains: "There is a practical issue in measuring data in the rotating frame - ie on the rotor - should that prove necessary, but that is less of a problem now than hitherto. There might be problems in measuring any new parameters identified - for example strain, but it is not possible to comment meaningfully until they have been identified."

Even so, it is a challenging problem. The CAA report comments: "The task is made extremely complicated by the fact that the aerodynamics of the rotor, and hence the details of its loading, vary considerably with flight condition."

An equally fundamental issue for the CAA and its sister bodies is whether it is worth funding this line of research, given that the service record clearly shows there is a declining risk. For the UK, in 2005 the rotor faults per flight rate was only one-seventh of that recorded in 1982.

Faults peaked in the mid-1980s and have broadly become rarer, but that has to be balanced against their potentially catastrophic consequences.

Howson notes: "In the case of tail rotors, it is hoped that a worthwhile safety benefit can be achieved with little more than improved analysis of data already collected - ie at minimal cost. The same cannot be said of main rotors, however, hence any future research would be conducted at a level commensurate with the size of the problem and the probability of success."

In practice that means the next stage of research may be a university-based study.

At the same time work is continuing with the participation of Bristow Helicopters and Meggitt Avionics to apply "anomaly detection techniques" to existing tail rotor data in the hope of demonstrating a usable defect-detection capability.

Howson says: "Completion of this work may also provide better insight into how best to tackle main rotor health monitoring. Consideration is being given to an academic study aimed at identifying the most promising means of detecting main rotor defects."


SOME KEY INCIDENTS INVOLVING ROTORS

  • March 1981 Sikorsky S-76A lost in cruise due to a main rotor blade loss following a spindle failure. Data lacking, but believed that detection of fault could have been possible.
  • July 1987 Eurocopter Super Puma forced landing with severe control difficulties due to detached frequency adaptor in rotor head following bearing wear. Unclear whether it could have been detected.
  • July 2002 Canadian military Bell 412 crashed fatally following a fatigue-induced tail rotor failure - one of seven related incidents. Believed likely that this could have been detected.
  • July 2002 All 11 occupants died when a S-76A lost a main rotor blade over the sea due to a fatigue crack following a repair to a blade struck by lightning. Researchers believe new HUMS techniques might have detected the crack.
  • June 2005 Super Puma ditched at sea following broken arm on tail rotor pitch-change spider. Data suggests that this might have been detectable with appropriate techniques.
  • October 2006 Main rotor spindle on a Super Puma failed due to fatigue crack as pilot attempted to lift off. A close call that caused great concern, but it is unclear whether it could have been detected.