I’m amazed a thread about HUMS could get past 4 or 5 replies without the Sultan jumping in, but let me throw in some fodder.....
HUMS is a great idea – with a bad execution.... HUMS could cause the most significant change in the cost of operating a helicopter that has occurred in decades.... HUMS is a smoke detector in a concrete room where you can’t store flammable objects.... HUMS is aviation’s answer to the chicken and egg paradox... HUMS is the bastard step-child resulting from a drunken night spent between a dynamicist and an avionics engineer. (or was that the Sultan). Anyway... All this is true to a degree,
The HUMS concept is simple. Lifed components (those which are replaced based on hours) have their life expectancy determined by assuming a conservative-case operating spectrum in a worst-case certified environment, plus a safety factor. Lifed components are that way because they typically fail in a way that is not preceded by some other effect (ie: cracks or wear indications that would otherwise indicate it’s time for replacement.) When they go, it’s sudden and nasty. So there has to be a good margin built into the calculation of life expectancy. The resulting schedule of required replacements and inspections based on flight hours is a major factor in the current expense of operating a helicopter – and it’s nasty.
So the concept of HUMS comes along with the promise to change all that. HUMS has really two parts – the logging part and the predicting part. They are both intended to allow us to extend the life of these components.
For the logging part, the HUMS logs the flight “regime.” It uses the flight sensors to accurately record how the helicopter is being used and for how long in each regime: Altitude, temperatures, torque, accelerations, and other information is recorded. What is recorded depends on what factors drive the life of the component (Is it altitude, temperature, hard landings, accumulated vibrational stress, etc.) This data is used in order that we can get away from the worst case assumptions about the flight regime. And this then allows us to extend component lives accordingly.
The other part is vibration trend data. As noted, if component failures were preceded by a predictable property, and that property exhibits itself well in advance of the actual failure, then we can use detection of that to replace components, rather than just clocking hours. In engines, the relative relationship between parameters (temperature / Ng, etc.) in the context of the flight regime, will give early indications of engine wear and impending failure without the need for adding sensors – just software. Helicopters usually require the addition of sensors. For example, certain bearings show a temperature rise over ambient prior to failure, others show vibration. Vibration is the most typical thing used to detect something starting to go wrong. (For years the HUMS was a 6th sense embedded in the butt of an experienced pilot.)
So the engineer in us says “THIS IS GREAT!!!” I can extend the certified life on components based on actual usage data, and I can reduce the margins for manufacturing variations, since if the failure does occur prematurely, my HUMS will detect it. Costs will go down... sales will go up.... avionics and airframe manufacturers will rejoice.... overall safety will go up.... there will be world peace...... (And I get a free RT&B system at the same time.)
So now comes the chicken and egg stuff. Enter the business folks and regulatory agencies (plus the product liability lawyers and all their friends). The Avionics guy wants to sell the system and make money. The operator needs the extended life credit in order to justify the expense of the system. The Helicopter OEM – who typically doesn’t make money on avionics – can’t justify the expense of determining the credit to be given. There is a huge expense to determining exactly how much the life can be extended, and how dependable the HUMS prediction is. The cost of that engineering, and the HUMS is going to drive up the cost of the helicopter. For his trouble, he’s going to lose on support sales. So unless you can show HUMS as directly contributing to selling more helicopters, its hard to get the Helicopter OEM to play ball. Add to that the potential for product liability – what happens on the day you lose an aircraft because the HUMS failed to detect an impending failure? How do you explain HUMS to a jury. So you either have to get a lot of test data, or a lot of in field data. For the in-field data to be significant, it has to go beyond the standard life of components.
Quite frankly, the best way for HUMS to really fulfill its promise is for it to become standard equipment delivered by the OEM and be the basis for component replacements used in the certification basis. That way everyone gets what they want. But the OEM has to be convinced that this provides the helicopter with a significant advantage in the market.
What has actually happened has been a bit of a ruse. Avionics manufacturers – who have the most to gain initially - have campaigned to the operators to demand HUMs, even when there is no component credit. The argument used to promote HUMs is safety. It has been made successfully enough that HUMS is now included as required equipment for some operators. In examination – that is a very hollow argument, but one that still gets made.
Advocates argue that the HUMS vibration monitoring allows the prediction of impending failures..... but remember how this works: HUMs detects failures which follow a predictable pattern of failure. By definition, if its predictable, then it should have been factored into determining the certified life of the component, and with the current margins, you are sure to replace the component long before you get into the pattern the HUMS was designed for.
The only safety improvement occurs IF the engineers were wrong in their calculations about the life of a component (applicable more to new designs), or IF a component fails – due to manufacturing defect - before its time, and then IF that failure happens to follow the pattern the HUMS was designed to detect, and IF the failure progress is still slow enough to still allow safe flight and landing. Then, man! it is your lucky day. That’s a lot of IF’s for a system that typically runs in the 6 figures.
Personally, I am an advocate of HUMS - but not the way I see it used today. I think HUMS has great promise, but I think operators are being duped into demanding and buying HUMS in a way that provides them very little for their money. I already know of some big operators who have spent millions on HUMS equipment to find it of little more benefit than a rotor track and balance system. As a result, I see potential for backlash before HUMS gets fielded properly. When that sort of thing happens, it will poison the market for the real HUMS with real credit, and real safety improvement / cost reduction potential.