DRAFT

Prognosis of Aircraft and Space Devices, Components, and Systems

Discovery Challenge Thrust (DCT)

Background: The USAF missions in air and space rely on the reliability of complex systems that range from aircraft and space platforms to electronic devices and sensors. Materials in these systems include a wide variety of metals, composites, polymers, and ceramics and combinations thereof ranging in forms from nanoscale quantum structures, to films and coatings, to complex structural components and structural assemblies. It is critical that these systems perform as designed for extended periods of time (much beyond their original design life). Past design practices have relied on various methodologies from safe-life to damage tolerance to reliability-based metrics such as mean time between failures (MTBF) for electronic components. Original predictions of performance for in-service applications have often been inadequate resulting in high costs for maintenance and repair, lack of availability or readiness, and in some cases loss of crew. Responses to these short-comings include in-service inspection requirements such as for aircraft structural integrity (ASIP), mandated corrosion inspection and repairs, line replacement unit upgrades in avionics, and various reliability improvement programs.

In no case are we able to predict by component serial number or other “finger print” identification of a device or component in a complex system when that device or component is reaching a state where it must be replaced or repaired. Instead, for systems we rely on system or fleet averages driven by the statistics of the lower tail of the reliability distribution. The USAF relies on statistically variable inspection methods for components that require large levels of damage in the form of cracks or visible corrosion in order to have a reasonable likelihood of finding the damage before component and system failure occurs. Great expense to the USAF occurs as a result of unnecessary and damaging inspections driven by worst-case conditions. Repetitive inspections are required to give a needed level of confidence that the damage state has not been missed.

This methodology and mind-set has driven the entire field of component and system reliability for non-electronics to focus the research on end-of-life scenarios – large cracks and extensive corrosion, for example. In the field of electronics, line replaceable unit (LRU) actions are driven by an assumption that all LRUs behave at the level of the worst case. Space electronics do not even allow for replacements! Reusable space access platforms such as the national space transportation system orbiter will end up requiring extensive ground time between missions to assure reliably the ability to launch the next mission for that platform.

The USAF needs the ability to design and deploy devices, components, and systems based on: (1) the ability to accurately predict with (defined) confidence the performance by serial number or other identifying “finger print”; (2) accurate prediction with defined confidence of future performance capability and potential degradation in near real-time; (3) actionable information to the operator so that corrective actions can be taken in a timely manner to assure mission completion and minimize operating cost and risk.

Problem: Determine in real time the current state of a uniquely-identified device, component or system with a defined degree of confidence such that the remaining capabilities of that system or component can be predicted with a high degree of accuracy and known level of confidence for any material systems or material combinations, material processing, operational environments, component usage history, and failure or material/structure/system degradation mode.

Technology Challenges: In view of the above problem and realities, the following taxonomy of critical future technology challenges has been identified:

1. Determine and assess early and progressive changes in material state associated with operational usage and exposure (again, any material and scale appropriate to the various damage states that must be captured).
1. Since the final failure location in any device, component, or system is most likely to be unobserved, any local state sensing must be able to map across the entire device, component, or system based on a suitable model thereof for all potential failure states for any internal and interface locations.
2. Fundamental innovations in traditional non-destructive evaluation for state awareness are required. Physics-based material-state sensing including novel signal processing technology applicable to unique mechanism detection and characterization through state change detection that include damage detection, material and damage state characterization need to be developed.
3. All state sensing or characterization models must provide actionable interpretation in terms of device, component, or system capability in near real-time such that combinations and synergisms of state changes are fully captured.
4. Strategies are needed to define optimum combinations of state awareness sensing and virtual (computational model-based) “sensing” are needed to define integrated system health management architectures.
2. Accurately predict the real-time physical, chemical or electronic state at any location for complex systems subject to hysteresis, damping, degradation, cyclic loading, thermal and environmental exposure over time, the predictions leading to state awareness.
1. State awareness and capability prognosis at any level must be able to be queried as needed for device, component, or system future capacity and reliability.
2. State awareness must include the ability to determine with defined confidence measures the existence of “hot spots” of damage or deterioration based on remote measures.
3. Modeling systems must be capable of multi-scale nonlinear representations of all state condition and processes and allow for full validation strategy development and demonstration.
3. Relate the current and evolving state of microstructure and damage processes at various length scales that will enable comprehensive probabilistic prognosis modeling of the material/structural/system state. Consider the following as a partial list of variables for state modeling:
1. Role of surface and interface material state changes and interactions associated with machining, surface treatments, films, coatings, cyclic hardening and softening, and environmental interactions critical to device, component, and system response and reliability.
2. Depending on the material system and operational-environment conditions, gradients and linkages across materials and interfaces, time and physical scales, physical and chemical processes, and other local variables and conditions critical to mechanism-based modeling at the device and component level could be considered,.
3. Uncertainty and variability effects must link to the physics/chemistry and mechanics in order to accurately account for response variability effecting device, component and system reliability and model confidence.
4. Processing, state evolution modeling and remaining capability prognosis must be provided in an end-to-end linkage to operational usage and environmental exposures such that the critical, physics-based parameters can be linked back to the original material processing conditions.
5. Interactions and synergisms between processing and intrinsic material state must be accounted for at the appropriate scales based on physics-based modeling.
6. Ability of devices, components, and systems to be aware of various regeneration, recover, heal, etc. and to communicate and participate in autonomic behavior is needed to assure the capability for fulfilling the original mission in the presence of damage or degradation.

Ability to reason through discontinuous past experiences including repair, autonomic response, and other system influences in the presence of current state awareness is needed in order to reliably predict future states with defined confidence measure at the device, component, or system level.