On-line Monitoring System For Structural Safety Based On Fatigue And Fracture

It is very important to ensure the safety for large-scale complex structures such as aircrafts. In the stage of design, the service lives and safeties of structures are predicted and designed by the fatigue and fracture assessments for various given loading conditions, respectively. However, the deviations between the given loadings and the actual service loadings are unavoidable and will cause potential danger for structures in services. For bridging the gap of lives and safeties from design to service, it is necessary to monitor the loading conditions of structures in real time. Moreover, the damage developments and residual service lives will be predicted by fatigue and fracture analyses using the monitored loadings and can be presented online. Here we proposed a structural safety guarantee system using the testing technology, signal processing technology and Ethernet technology based on the three-dimensional fracture mechanics theories. The main contributions are as follows:(1) A kind of structural safety on-line monitoring system is invented. For the safety problem caused by the differences between the actual service loadings and the assumed loadings used in the stage of design, a kind of structural safety on-line monitoring system base on fatigue and fracture theory is invented. It can monitor the actual loadings endured by the structures in service and analyze the propagations of the damages on line based on the monitored loadings. When the predicted damage reaches the dangerous level, the system raises the alarms. The system can eliminate the potential safety hazard caused in the stage of design and improves the safety of important systems such as airplanes.(2) A kind of data acquisition and processing module base on hardware protocol stack is developed. The classical approach to realize the Ethernet communication is based on software protocol stack, which occupies a large amount of resources of the processor. But the on-line monitoring system needs to process the complicate algorithms in real time. For the problem, the hardware protocol stack based Ethernet communication technology is applied in the on-line monitoring system successfully, which solved the contradiction between the communication and real time processing and improved the stability. What is more, it also saved the precious storage resource of the on-line monitoring system. Because the system has the Ethernet interface, it can be used to construct the network monitoring system very easily.(3) A kind of strain gauge monitoring unit is invented. For the on-line predictions of the damages conducted by the on-line monitoring system are based on the actual loadings monitored, the strain gauge's failure will cause a series of wrong prediction results, which may cause serious accident. For the problem, a kind of strain gauge on-line monitoring unit is developed. It can be connected to the strain-measuring bridges to monitor the working state of them, with on effect on the monitoring results of the structural safety on-line monitoring system. When any strain gauge of the bridge fails, the unit can raise alarm and point out the failed strain gauge, so as to ensure the reliability of the monitoring results. When the monitoring scale is large, it can also obviously reduce the the working loads of checking the strain-measuring bridges.(4) A new overload retardation model based on 3D constraint factor is developed. Because of the strong relation between the precision of the on-line monitoring system and the fatigue crack propagation models, the studies on the crack propagation is conducted and an overload retardation model suitable for the on-line monitoring system is developed. When the crack growth rate at constant amplitude loading is obtained, the overload retardation factor of this model can be determined by a retardation test under any overload ratio. The retardation effects of other overload ratios can be predicted by the obtained factor. Base on the factor, the developed model is proved to be efficient on aluminum alloys such as 2024-T3, 2024-T351, 6061-T6. In most cases, the errors of lives between prediction and test are within 20%. The errors are still less than 34% even for large-scale yielding cases. More importantly, it is also shown from the verification for 2024-T351 that the new model is effective for predicting the thickness effect on overloading retardation.