Downloaded from https://www.everyspec.com ADS-79D-HDBK M.6.3.1 Data Validity: The validity of condition data is dependent on multiple factors. One of the primary factors is proper tracking of the data to the component serial number. Other factors include maintenance actions and operational and environmental conditions, which should be monitored and documented. The data validity of the condition monitoring tools should also be monitored. A few examples are provided: M.6.3.1.1 Off-aircraft: For laboratory and at-line instruments, standardizing daily ensures the validity of data.
These standardization checks may be internal to the instrument or may utilize actual samples similar to the oils being tested. If the daily standardization fails, the instrument should not be used until the cause of failure is determined, remedied and standardization passes. See Section M.6.1.1.3. M.6.3.1.2 On-aircraft: For on-line and in-line sensors, the data acquisition system should have a built-in test capability to check the validity of the incoming sensor data. 93 The system should notify maintenance personnel if the sensor is suspected of being faulty. The integrity of a monitoring system should be ensured by having the ability to detect and flag faulty sensor data.
When faulty data are detected, a process should be in place to account for the lost information. Linear interpolation between good values is typically used. 94 M.6.3.2. AOAP: AOAP data is automatically uploaded to Oil Analysis Standard Interservice System (OASIS) for most instruments and OASIS is automatically updated to LIW for easy access by all units. M.6.3.2.1Chip Detectors and Mesh Screens: There is generally a light on the maintenance panel indicating when these devices detect debris. Any visual debris observed during phase inspections is manually annotated in a logbook. The introduction of any new system should be at least as good as the current process.
At- line instruments should aim to have their data uploaded to the ground station and on-line sensors should aim to have their data automatically uploaded to the on-board portion of the HUMS. M.6.3.3 Data Development: Statistical process control (SPC) and cumulative distribution techniques are generally used to statistically evaluate alarm limit values for oil analysis. The data set utilized should represent a normal distribution where the majority of samples are expected to fall within two standard deviations of the mean or represent about 94% of all samples taken. Note some oil measurements are bound by zero and the distribution will not appear normal.
While a mean and standard deviation can still be calculated, the user should verify that alarm limits based on these statistics are descriptive of the actual distribution. For 93 ASTM D7685-11, Standard Practice for In-Line, Full Flow, Inductive Sensor for Ferromagnetic and Non- ferromagnetic Wear Debris Determination and Diagnostics for Aero Derivative and Aircraft Gas Turbine Engine Bearings. 94 ASTM D7669-11, Standard Guide for Practical Condition Data Trend Analysis. 256