The Simple Question
It’s a simple question, and for today’s ditty I am only going to consider aircraft where the maintenance manual limitation is expressed in normal acceleration units (g). There are a number of aircraft (Embraer types and BAE 146 for example) which express hard landing criteria in vertical speed at touchdown, and this blog does not address those aircraft.
Somewhere in the maintenance manual there will be a limit such as “If the acceleration at touchdown exceeds 2g, perform a hard landing maintenance inspection”. We have an apparently simple question to answer, namely did the normal acceleration at touchdown exceed 2g?
Life is not that simple
Of course, as readers of this blog will already know, life is not as simple as it appears. The problem in this case is that accelerometers are not perfect, and can suffer from offsets. Let’s do the sums with a nominal 0.1g offset.
If the accelerometer has a +0.1g offset, it will read 1.1g when sitting still on the ground (OK, for the pedants, I am assuming that the aircraft is both wings level and at zero degrees pitch at a location on the earth with a perfect 1g local acceleration due to gravity. We are not catering for an aircraft sitting with a nose up attitude, one wing low due to sloping tarmac and over a huge underground cave, all of which would reduce the measured acceleration).
Back to our aircraft which lands with a true bump of 1.95g. This will be measured by our inaccurate accelerometer as 2.05g (remember, +0.1g offset). The options are now:
- Declare this to be a hard landing because the measured acceleration exceeded the maintenance manual limit by 0.05g, or
- Declare this NOT to be a hard landing because the corrected normal acceleration was 0.05 g below the threshold.
Of course, there is the other case:
If the accelerometer has a -0.1g offset, it will read 0.9g when sitting still on the ground. Our aircraft lands with a true bump of 2.05g. This will be measured by our inaccurate accelerometer as 1.95g and the options are:
- Declare this to NOT be a hard landing because the measured acceleration is below the maintenance manual limit, or
- Declare this to be a hard landing because the corrected normal acceleration was 0.05 g above the threshold.
…it gets worse…
Some operators we know of process data using two different analysis systems, and with one applying a correction while the other does not. They then have one system calling for a maintenance inspection and the other does not, and this simply reduces confidence in the analysis process. After all, you would expect any system to give an alert in such a simple case.
How bad is it?
The problem would not exist if the accelerometers were perfect. Equally, if the accelerometer offset was tiny, it would very rarely arise. So how big can the offset be?
Turning to ED-112A, the range is ± 0.09 g excluding a datum error of ±0.45 g. It’s a staggeringly large offset, and pretty much makes a nonsense of using this signal without correction for measuring hard landings. In practice, it is not uncommon to find accelerometers with offsets of greater than 0.1g, although I have never seen one greater than 0.2g.
Notes for Maintainers
If you are a maintenance engineer working with an uncorrected system that reports a hard landing, I suggest you check the value where the aircraft is sitting still at the end of the data and, if it has a significant offset, subtract that from the peak to see if a maintenance inspection is really called for. You may save yourself effort, and get the aircraft back in the air sooner.
Equally, if the offset is in the opposite sense, there may be hard landings that you are not being alerted to. A case of ignorance is bliss?