Downloaded from https://www.everyspec.com ADS-79D-HDBK Retirement methods based on load monitoring should be documented in detail in the CBM Management Plan. Reliability analysis of the resulting retirement scheme should be based on statistical evaluation of fatigue strength. (see section A.6.3 of this ADS for additional reliability guidance). C.5.4 Load estimation. Load estimation, which is also known as ―virtual loads modeling‖, is a method of processing aircraft state and control data per rotor cycle and estimating maximum, minimum, vibratory, and steady loads. If possible, the variance in each aspect of load should also be estimated.
Various models may be formulated, including purely phenomenological, purely statistical, empirical, or neural-network based. Model validation should include comparisons to available flight test loads per rotor cycle. However, validation of statistical, empirical, or neural-network models should not rely on the same test data used to develop, or train, the model. Diagnostics tools for load estimation should incorporate diagnosis of the repeatability and accuracy of the reported state and control data used by the model.
In addition, exceedances of the tested state and controls envelope used as a basis for the load estimation models should be indicated to allow for predetermined special maintenance procedures, as appropriate. Retirement methods based on load estimation should be explained in detail in the CBM Management Plan. Reliability analysis of the resulting retirement scheme should be based on statistical evaluation of load estimation and fatigue strength. C.5.5 Limitations on use of vibration measurements for structural health monitoring.
Vibration measurements, including processed signals such as vibration based ―condition indicators‖ or ―health indicators,‖ may provide an indication of structural damage, such as rod end bearing free play and airframe cracks. However, the vibration frequencies and load magnitudes measured in airframe structure during flight are directly related to applied loads. The most appropriate use of vibration measurements as an indication of structural deterioration may be to monitor changes in natural frequencies due to stiffness changes. This approach would require periodic dynamic inputs and the measurement of frequency response.
However, this approach would not be suitable for structures which exhibit an immeasurable change in dynamic response prior to primary structural load paths being severed, including non-redundant structure and any structure prone to wide-spread fatigue damage. Partially through cracks will not produce a significant change in structural load paths or stiffness. Hence, vibration measurement may not be sensitive to structural deterioration and are considered a poor indicator of structural health for airframe structure. Therefore, when any such vibratory indication of structural damage is provided, the structure near the vibratory indication should be inspected using established NDI methods.
Prior to burdening maintenance personnel with required but potentially unnecessary inspections, the physics of structural health should be clearly linked to load path and vibration indication. Vibration measurements are not considered a viable approach for achieving airworthiness credits for use in place of scheduled airframe structural inspections. 70