N211-014 TITLE: Predictive Model Based Control System for High Speed Dynamic Airframe Testing
RT&L FOCUS AREA(S): General Warfighting Requirements
TECHNOLOGY AREA(S): Air Platforms
OBJECTIVE: Develop a scalable, real-time, predictive, and adaptive model-based test frame control system that increases load cycling frequency while maintaining load accuracy for high speed dynamic rotary wing airframe testing.
DESCRIPTION: Full-scale fatigue testing is required for all new aircraft designs. While all aircraft are subjected to this testing, rotary wing aircraft often prove to be much more difficult to evaluate because of the high cycle counts that helicopter airframes experience. Currently, the limitations of structural testing control systems require full-scale fatigue tests to be performed at frequencies much lower than those generally experienced on rotary wing aircraft. Full-scale fatigue testing on rotary wing aircraft is typically limited to a low cycle fatigue test, where Ground-Air-Ground cycles and simplified maneuver loads are applied to the airframe. Truncation and/or equivalent damage methods are used to reduce the cycle count in order to perform a test within a reasonable time period. However, there is evidence that shows that equivalent damage methods, which remove high frequency load components at high mean stress loads, can produce unconservative crack growth rates. The crack growth rates are slower than what would be accumulated on an in-service aircraft, which creates a risk of not being able to find premature cracking at a representative time, or even at all during the full-scale fatigue test. Since pure cycle count reduction cannot produce test results that are consistent with real fleet usage, increasing testing speed is required to be able to incorporate more loading cycles without significantly prolonging a test effort.
Current technology used in full-scale fatigue testing is limited to load cycle speeds of approximately 2 Hz, and most tests are practically slow enough to be considered quasi-static. At speeds this low, full-scale fatigue tests would take over 200 years to complete if all vibratory load content were to be included. The control system generally used for this testing is a reactive-style feedback loop that requires a load to be applied, usually by means of hydraulic servo-cylinders, and the system response to be read by sensors, such as strain gauges and load cells. Gains in the feedback loop are adjusted to provide satisfactory tracking between the target and measured loads or strains. While these reactive methods are adequate for quasi-static tests, they become insufficient as the frequency and speed of the test increases due to complexities caused by large airframe displacements, airframe inertial effects, actuator cross coupling, and phase lag caused by system response times. If these issues are unaddressed, the load cycling rate in a test will have to remain low in order for loads to be applied accurately. Accurate loads are required to attain representative test results to ultimately make a correct assessment of the actual life of the airframe, as well to catch and predict early cracking that might occur in the fleet.
This SBIR topic seeks a model-based, or "model-in-the-loop", control system for full-scale aircraft fatigue testing that can achieve higher cycling rates and faster test speeds compared to those achievable by current reactive control systems (0.5 Hz � 2 Hz). The control-system should be able to predict and generate the signals required for load application based on sensor data (including strain gauge bridges, load cells, displacement transducers) and a representative model of the system. This model could include the test article, fixtures, actuators, hydraulic valves and supply system, and sensors located on the test article or on the actuators. A peak loading frequency of at least 10 Hz is desired in order to match the primary loading frequencies on rotary-wing platforms. The control system should be capable of controlling high speed actuators that can achieve speeds in excess of 100 in/s in order to meet or exceed the frequency requirement while still being able to achieve displacements that may be several inches in magnitude. The controller should be able to simulate the test system in real time, use the model to predict required actuation signals, adapt the model and parameters to account for nonlinearities and uncertainties, and be scalable to handle multiple degrees of freedom with coupled actuations with potential for 15 or more actuators.
Commercial and naval aircraft both face similar requirements for full scale fatigue testing. Improvements to testing speed while maintaining required loads and displacements would improve both cost and schedule for acquisitions and validation of new platforms. This technology could also improve dynamic testing in automotive applications, as well as for other ground-based military vehicles.
PHASE I: Determine feasibility of a real-time, predictive, and adaptive control system using a simplified test setup that leverages models of the test system in order to increase variable amplitude load accuracy at higher frequencies. Develop a plan for expanding the Phase I work into a prototype system that can be demonstrated on a simplified test article capable of increased test speeds and controlling multiple actuators.
PHASE II: Develop and demonstrate a prototype model-based control system based on the Phase I approach by integrating the controller into a test that applies representative loads onto simplified test article that is representative of an airframe structure in order to show increased control system performance (i.e., speed and accuracy) against a traditional control system. Demonstrate the ability to handle the coupling of multiple actuators as seen in a full scale fatigue test.
PHASE III DUAL USE APPLICATIONS: Develop and demonstrate a modular and scalable model-based control system on a full scale fatigue test specimen using multiple actuators and combined vibratory/maneuver loading. Verify that the system can apply vibratory loads accurately at load cycling rates of 10 Hz or higher.
Because commercial and naval aircraft both face similar requirements for full scale fatigue testing, improvements to testing speed while maintaining required loads and displacements would improve both cost and schedule for acquisitions and validation of new platforms. This technology could also improve dynamic testing in automotive applications and for other ground-based military vehicles.
REFERENCES:
KEYWORDS: Control; Dynamic; Testing; Structure; Predictive; Adaptive