Research On The Reliability Test Of Complex Mechanism Of Aircraft Considering Multi-factors

Complex mechanism exists in many modern aircrafts, and it affects directly theflying safety and mission capability of aircraft. As an effective method to excite thepotential failure modes, and to improve the mechanism reliability, the reliability testis indispensable in the designing process of aircraft.The failure mechanism of aircraft mechanism is affected by both working loadand combined environmental stress, and the wear of kinematic pair exists only in themotion process of mechanism. Consequently, in order to excite the potential failuremodes of aircraft mechanism, the multi-factors, including working load andcombined environmental stress and motion function, should be considered in thereliability test of aircraft mechanism.The traditional reliability test methods for mechanism are based on the statisticsof failure event, and often consider very few factors. Consequently, they will notwork well when the aircraft mechanism has very high reliability and very longlifetime. In addition, as an important method to get the dispersity information ofmechanism lifetime, the simulative reliability test methods of the moment considervery few factors, and don’t consider the coupling mechanism of factors. As thebasement of mechanism optimization, the present IM (Importance Measure) methodsof random variables can not represent the contribution of the distribution parametershonestly. Consequently, based on the discussion above, the detailed contents are asfollows:(1) A reliability test method considering multi-factors is proposed for complexmechanism of aircraft. In the proposed method, the reliability test and analysis formechanism components will be executed firstly, where the multi-factors areconsidered. Secondly, the reliability test for mechanism system will be executed,where the key factors are considered. In the test process, the reliability test methodwith lifetime as characteristic quantity is used to assess the reliability of mechanism,and the reliability test method based on safety boundary is used to study the failuremechanism. The two methods mentioned above are supported by each other, andsynthesized into an integrated system. In addition, the key technologies of the testmethod are studied also. (2) For the purpose of reliability assessment, a reliability test method withlifetime as characteristic quantity is developed based on the present methods, and thereliability assessment method is derived. The test method is executed as follows:firstly, the sample size and test time are derived based on the design lifetime, and thedispersity information of lifetime, and confidence coefficient. Secondly, the testprogram is designed and then the test is executed. If all the samples can finish thetest with out failure, the reliability performance can be regarded to fulfill the designrequirements. If not, the mechanism reliability will be assessed, and the failuremechanism will be studied using the test method based on safety boundary.(3) In order to study the failure mechanism of failure modes, a reliability testmethod based on safety boundary is proposed. The test method is executed asfollows: firstly, the failure modes are derived based on empirical analysis, and thefailure mechanism are studied using the physical failure simulation test. The keyfailure modes and key factors can also be derived based on the test results. Secondly,the safety boundaries of the key failure modes can be derived, with the physicalfailure simulation test as main method and the numerical simulation as assistance. Atlast, the reliability can be assessed based on the safety boundary and the jointPDF(Probability Distribution Function) of the key factors, and the improvementplans can also be proposed based on the test and analysis results.(4) In order to get the dispersity information of lifetime, a simulative reliabilitytest method considering coupling mechanism of multi-factors is proposed. In theproposed method, the factors of mechanism are classified as original factors,cumulative service factors and non-cumulative service factors, and the situation areclassified as original situation and working situation. In the simulation progress, thereliability model is constructed firstly and the simulation samples are derived usingMont Carlo method secondly. Thirdly, taking the motion cycle as simulation step, thevalue of cumulative service factors are computed based on the working condition ofthe prior step, and the output value of the reliability model is computed step by step,until all the samples failed. Finally, the reliability evolution process and thedispersity information of lifetime can be derived based on the simulation resultsabove. In the last, the reliability evolution process of the gear door locks are studiedusing the method proposed above, and the dispersity information of their lifetime arederived.(5) The test systems and test devices considering multi-factors for the gear door locks and the lock systems of an aircraft are designed respectively. As the core partof the test systems, the test devices can realize the simulation of motion function andthe loading of aerodynamic force for the locks and lock systems. Moreover, theweight and the stiffness of the test devices fulfill the control requirement of vibrationplatform, and their vibration performance fulfills the test requirement. Consequently,with the assistance of other test equipments, such as the vibration platform, and thecombined test chamber, et al, the multi-factors (including working load, andcombined environmental stress, and simulation of motion function) can be loaded.(6) Taking the gear door lock systems as example, and based on the test systemsand the dispersity information of lifetime derived above, the engineering applicationof the test method proposed in this paper is illustrated. The test results show that thereliability and lifetime of the gear door locks don’t fulfill the design requirement, andbased on the failure mechanism analysis, the improvement plans for the locks andthe lock systems are proposed. From the subsequent reliability test and actual serviceperformance of the lock systems one can see that the improvement plans areeffective, and fulfill the design requirements.(7) Taking the slat mechanism of an aircraft as example, the loading program ofthe aerodynamic force is studied, and the servo loading system is designed. In theloading system, the simulation forces are applied on the slat surface by lever linkagesand wire ropes, and the force direction is controlled by slide mechanism that derivedby servo motor, and the force value is controlled by DC(Direct Current) motor.Consequently, the equivalent action point, and the value, and the direction of thesimulation force will be equivalent with the actual aerodynamic force. The guiderails of the adjacent loading points are designed together in order to decrease theloading system size, and they are placed in slope in order to decrease the guide raillength. The slider wheels are designed with rim so as to undertake the horizontalforce. In addition, the loading system of the wing is designed based on localcoordinate system, where the deformation of the wing is applied by motor derivedscrew elevator. In the last, the test programs for the slat mechanism are discussedbriefly.(8) A new IM (Importance Measure) method based on CDF (CumulativeDistribution Function) is proposed. One linear and two nonlinear models areintroduced firstly, and the distribution parameters of random variables are classifiedas “shape parameters”(such as the variance of normal distribution) and “location parameters”(such as the mean value of normal distribution). Based on the discussionabove, the applicability of the moment-independent IM method is studied, and onecan get that the moment-independent IM method will not represents the contributionof the location parameters to the importance values honestly. In order to eliminatethis defect, a new IM method is developed from the moment-independent method byreplacing the unconditional PDF with the conditional CDF under the condition ofXi=0, and replacing the conditional PDF with the conditional CDF. Discussion andanalysis shows that the new IM method has the properties of global, and quantitative,and model free, and moment-independent. Emulation results of analytical test casesand application example show that the new IM method can represents thecontribution of all the parameters to the importance values honestly.