| Thrust weight ratio is increasing and thin casing structure is widely used in the modern aero-engine,which has a great effect on the dynamic characteristics of the whole engine for the coupled vibration between rotor and stator.The dynamic characteristics mainly include critical speed analysis,whole system's response mechanism and mode coordination between rotor and stator.It is significant to improve the safety and reliability of aero-engine by controlling the vibration of the whole engine and optimizing engine structure in the design stage.Some methods can be provided to control vibration,such as studying the mechanism of rotor-stator coupled vibration,building a more scientific quantitative evaluation index of whole engine vibration,developing advanced and efficient structural optimization methods by combining modern machine learning methods and artificial intelligence technology.This paper focuses on the topics of coupled vibration of whole aeroengine structure system and its intelligent optimization.The main contents are as follows:1)The finite element model of an aircraft engine rotor tester with casing was built,and the modal test of the whole tester was conducted.Due to the asymmetrical stiffness in the horizontal and vertical directions,the modal test of the entire tester was performed in two directions.Based on the modal test results,the support stiffness and the mounting stiffness of the tester were identified intelligently by the method,combining the support vector machine and genetic algorithm.Using the identified stiffness,the harmonic response of the tester was simulated in the horizontal and vertical directions,and compared with the measured frequency response of the testing points.Each testing point had been achieved satisfactory consistency,which verified the effectiveness of the identification method.2)The single ring equivalent modeling method for aero-engine rotor blade model was improved,and a multi-rings equivalent method was proposed.The two-ring equivalent method was verified by a simple disk shaft system with blades.The blades were simplified by two-ring equivalent method.The modal analysis was performed on the original model and the simplified model.Then the results were compared with experiments.The free vibration modes of the first 4 orders(excluding the first 6 rigid body modes)were consistent,the natural frequency error was within 1%,and the errors of the first three critical speeds were within 0.5%.The number of elements was reduced by 34.8%,the number of nodes was reduced by 38.5%,and the calculation time was reduced by 44.8%.This method was applied to a certain large bypass ratio double-rotor turbofan engine,and the finite element model of a certain engine was established.Free modes of the high pressure rotor system,the low pressure rotor system,stator system and the dual rotors system which includes the high and the low pressure rotors,were analyzed.Critical speeds of the high and low pressure rotors without stator system were analyzed,and critical speeds of the whole engine including the mounting and the stator also were analyzed.The results showed that compared with the dual-rotor system,the order of critical speeds and the speed value of the whole engine had been changed.The mode shapes of the rotor system were approximately same.3)Based on the existing aero-engine structure design guidelines,a critical speed risk coefficient,a rotor strain energy risk coefficient,and a cross-section rotor-stator rubbing risk coefficient were proposed to evaluate quantitatively the coupled vibration of the aero-engine.Rationality of the three indicators was discussed.Using the proposed indexes,coupled vibration of the aero-engine rotor tester with casing and a certain large bypass ratio double-rotor turbofan engine were evaluated and analyzed.The calculation methods of these three coefficients were presented.Rationality and effectiveness of evaluated coefficients were verified.4)Based on the proposed three evaluation indexes,the mechanism of coupled vibration of the aero-engine rotor tester with casing was studied.The influence of the support stiffness and mounting stiffness on the first three critical speeds risk coefficients,the rotor strain energy risk coefficients,and the rotor-stator rubbing risk coefficients of the key cross-sections,were analyzed.The generation and variation factors of the static coupled vibration were explored.The analysis showed that the defined index parameters can reflect accurately the dangerous degree of the critical speeds of the tester,the strain energy of the rotor at rotor-stator rubbing cross sections.5)Based on the proposed three evaluation indexes,the whole engine coupled vibration mechanism of a certain large bypass ratio double-rotor turbofan engine was studied.The critical speed risk coefficients,the rotor strain energy risk coefficients,and the cross-section rotor-stator rubbing risk coefficients at five key sections,such as fan section and the first low-pressure compressor section,were analyzed.The influence of each support stiffness on the coupled vibration was investigated.The results showed that the impact of each support stiffness on each index was extremely complicated.The design of support stiffness was affected by many factors,and the mutual constraint relationship of each index needed to be satisfied.It was difficult to optimize the aero-engine support stiffness.6)A multi-objective intelligent optimization design method for aero-engine bearing stiffness was proposed.And the support stiffness of a certain large bypass ratio turbofan engine was optimized.Taking the support stiffness as the design variable,the sampling method was used to extract the sample stiffness combination in the stiffness variable space.Sample data of “stiffness-design index”were calculated.Using support vector machine regression to get a computational agent model of the "stiffness-design index",the NSGA-II(non-dominated sorting genetic algorithm-II)was used to optimize,and ‘Pareto' stiffness was obtained and filtered.Finally,the required design stiffness was obtained.The results showed that the coupled vibration of the whole engine can be optimized by this method.Three sets of results were chosen as the final optimization results.The first set of solutions reduced the critical speed risk coefficient by 5.79% and the rotor strain energy risk coefficient by9.36%,the cross-section rotor-stator rubbing risk coefficient was reduced by 8.6%.In the second group,the critical speed risk coefficient was decreased by 2.95%,the rotor strain energy risk coefficient was decreased by 13.12%,and the cross-section rotor-stator rubbing risk coefficient was reduced by 7.13%.In the third group,the critical speed risk coefficient decreased by 4.80%,the rotor strain energy risk coefficient reduced by 17.25%,and the cross-section rotor-stator rubbing risk coefficient decreased by 2.94%. |
PDF Full Text Request