Research On Impact Load Identification Technology Of Stiffened Composite Structure Using Case-Based Reasoning

Composite material has the advantages of high specific strength,high specific stiffness and the property to be designed and so on,which makes the composite structure of aircraft become the development trend of modern aircraft.In addition,the composite structure of aircraft is an important symbol of the advanced aircraft.Compared with metal materials,composite material has the poor performance of impact resistance.Moreover,the composite structure of aircraft may suffer from different types of external impacts(such as tool drop,hail impact,bird impact,etc.)during the process of manufacturing,service and maintenance,which may lead to barely visible impact damages(BVID)for composite structure.If these damages cannot be detected by the user or the maintainer in time,the bearing capacity of the composite structure will be reduced.In severe cases,it may even cause the structure sudden failure.Therefore,it is need to monitor the impact events of composite aircraft in real-time and on line for improving the safety,reliability and service life of composite structure and decreasing the maintenance cost.This paper is based on the research background of the project of an aircraft's composite empennage force-heat coupling ground test health monitoring,which has been undertaken by our research group.Then,a stiffened composite laminates that can simulate the typical structure of aircraft composite empennage and its corresponding impact monitoring system are constructed.A case-based reasoning(CBR)technology is proposed to identify external impact event for a stiffened composite laminates.Furthermore,the corresponding experimental verification is carried out.The main research work of this paper is summarized as follows:(1)The existing methods of impact load identification and their advantages and disadvantages are analysed and contrasted.Then according to the basic principle of CBR technology,the impact load identification method case-based reasoning is designed and the corresponding workflow is constructed.In addition,according to the research needs of the project,a stiffened composite laminates was designed and manufactured and some piezoelectric smart layers that can withstand up to 150 ? high temperature were manufactured.Moreover,the application stability performance of piezoelectric smart layer in the high temperature environment is verified.Finally,an impact monitoring platform based on piezoelectric smart layer for composite structure is constructed.(2)The propagation characteristics of stress wave that is caused by external impact in stiffened composite laminates are studied.The stress wave feature cased by the impactors with different elastic modulus,the attenuation characteristics of the stress wave in different directions and the relationship between the signal energy and the impact energy are analysed emphatically.The experiments results show as follows:(?)The higher the elastic modulus is,the more the high frequency component will be.(?)The propagation attenuation characteristics of stress waves are different in parallel and perpendicular to the stiffener.(?)The impact height is proportional to the signal energy and the signal energy is also proportional to the impact energy.(3)The construction and characterization method of impact case unit(ICU)in the technology of impact load identification case-based reasoning was systematically studied.The influence of feature type and feature number of the ICU components and sensor network on the similarity of ICU were emphatic researched.By comparing with the correlation coefficients of each ICU constructed by different types of features at the same location,the waveform feature after de-averaging and normalization was selected as the characteristic parameter to construct the impact location component and the global signal energy was selected as the characteristic parameter to construct the impact energy component.In addition,three different impact casebases based on different sensor network were constructed.Finally,by comparing with the mean and variance of the correlation coefficients of each ICU in the three impact casebases,the sensor network that makes the difference between each ICU in the casebase larger was selected,(4)The nearest neighbor algorithm was used to study the impact load identification case-based reasoning with the impact data of different height at the any ten points on the stiffened composite laminates.The location identification results show that the method of impact load identification case-based reasoning has better identification results.Besides,the impact energy identification results show whether the impact point occurs in the stiffened area or the non-stiffened area,as long as the most relevant intersection obtained by using the identification method case-based reasoning is also in the same area,the error of impact energy identification is within the acceptable range.