Abstract: In order to more comprehensively and effectively monitor the evolution process of vibration performance state of rolling bearing in service，the nonlinear and uncertainty features of evolution of vibration performance state were characterized by correlation dimension and box dimension，the fluctuation ranges of instantaneous value and average value of vibration data.The grey confidence level between the parameter sequences was solved by using the grey relation analysis method，so as to analyze the nonlinear correlation degree between the nonlinear and uncertain characteristics of the evolution of the bearing vibration performance state.Two cases showed that the correlation dimension and box dimension had a trend of decrease，and the fluctuation ranges of instantaneous value and average value had a trend of increase on the whole.Moreover，they had corresponding relationships at the discontinuity points in a sense.There were negative correlation relationships among nonlinear parameters and uncertain parameters.Besides，the grey confidence levels between sequences of uncertain parameters and correlation dimension were higher than the uncertainty parameters and box dimension.
Abstract: Due to the inconformity of assembly stress and thermal expansion transfiguration in the internal components of shaft system，the precision angular contact ball bearing could be applied into the combined loads consisting of axial load，radial load，and torque，which affected both the dynamic characteristics of bearing and the service performance of precision shaft system.According to the relationship of interaction and relative position between the rolling element and bearing track，a five-degree-of-freedom quasi-static model considering the effects of centrifugal force，gyroscopic torque and other nonlinear factors was established，and the calculation model of bearing stiffness was improved by the principle of step-by-step search method.The evolutionary regularity of dynamic characteristics for precision angular contact ball bearings was investigated in combined loads.The results indicated that the axial load could decrease the difference between ball-inner raceway contact angle and ball-outer raceway contact angle，and suppress the stiffness softening caused by the increase of speed.When the torque and radial force were applied separately，the fluctuation range of contact angle and contact force increased，and the bearing stiffness along the direction of loads also increased，while the stiffness decreased perpendicularly to this direction.In the condition of combined loads，the torque would restrain the fluctuation of load distribution due to the radial load，and improve the loading capacity and service life of precision ball bearing.
Abstract: In view of the problem that information loss and false information may occur during the coarse-graining process of multi-scale dispersion entropy （MDE），which make it difficult to extract bearing fault information comprehensively，a rolling bearing fault diagnosis method based on improved refined composite multi-scale normalized dispersion entropy （IRCMNDE） and nearest neighbor convex hull classification （NNCHC） was proposed.The refined composite multi-scale dispersion entropy （RCMDE） was introduced，and the average value in the coarse-graining process was replaced by the maximum value to represent the data segments information，which can overcome the shortcomings of the traditional coarse-graining process and highlight the fault characteristics.Through the normalization operation to reduce the influence of the selection of different parameters on the entropy value，IRCMNDE was acquired as feature samples； NNCHC was used to classify the feature samples to realize bearing fault diagnosis.Experimental results showed that the proposed method can achieve 98.98% fault identification accuracy.Compared with the methods based on MDE （fault identification accuracy was 95.99%） and RCMDE （fault identification accuracy was 97.60%），the proposed method can extract the fault feature information of rolling bearings more accurately and improve the accuracy of fault classification.
Abstract: In order to solve the problem of large amount of model parameters and calculations in the existing fault diagnosis methods based on convolutional neural networks，an improved SqueezeNet model was proposed to be applied to the fault diagnosis of helicopter rolling bearings.By drawing on the idea of the VGG16 model based on the classic SqueezeNet，the model used three sizes of convolution kernels instead of one size of convolution kernel，and realizes the increase of network capacity，enhancement of nonlinearity，reduction of network parameters amount under the same perceptual field conditions.To further reduce the amount of network parameters，the convolutional layer，pooling layer，Fire module and pooling layer were alternated to form the model feature extraction layer.While ensuring the feature extraction capability，the amount of network parameters was further reduced.The model faults diagnosis experiment was carried out through the bearing data onto the research group.The results showed that the diagnosis accuracy of the model reached 99.65%，which was comparable to the traditional convolutional neural network and the classic SqueezeNet model.The calculation amount and the parameter amount were reduced by about 6 times and 36 times.
Abstract: Several types of short centerline probes were designed without need of cable fixation.The characteristics of the measured values were studied by wind tunnel test.The results showed that the difference of Mach number root-mean-square deviation between short centerline probe and long centerline probe with different shape parameters was less than 0.000 3，the difference of Mach number correction was less than 0.000 8，and the measurement curve had no obvious change when the Mach number was below 0.95；There were great differences between short centerline probe and long centerline probe of different shape parameters When the Mach number was between 1.0 and 1.4，changing the shape parameters of short centerline probe or increasing probe length can reduce the interference of head to flow field to some extent.
Abstract: The fluid network analysis method has its unique advantages in nozzle film cooling problems，but it cannot predict the influence of inlet/outlet flow on the static pressure change along cooling air channel.In order to solve this problem，a type of element that can describe the static pressure change caused by inlet/outlet flow was introduced，and the variation of static pressure was quantified based on the lag angle of velocity at hole entrance.Results on verification examples showed that the improved method can correctly capture the static pressure trend along the cooling air channel and reduced the calculation error of cooling air distribution to less than 0.5%.The improved fluid network analysis method was coupled with computational fluid dynamics simulation of mainstream，and used to analyze cooling air distribution in nozzle.Compared with computational fluid dynamics simulation for whole system，the method had very low computation cost，meanwhile the difference between two methods was less than 1.2%.The application in the design of a nozzle showed that the method reflected the influence of cooling structure changes on cooling air distribution effectively.
Abstract: In view of the three bearing swivel nozzle （3BSN） for turbofan engines，numerical simulation method was used to study the spatial distribution of infrared radiation signature of three bearing swivel nozzle under the working conditions of vertical take-off and landing （VTOL） and cruise（non-vectorial and 90° vectorial conditions），and the influence mechanism was also analyzed.The results showed that under non-vector state，the local high temperature area appeared on the lower wall of 3BSN due to the special geometry surface，and the jet shape became an elliptical cone，resulting in a maximum increase of 44.6% in wall infrared radiation from positive detection angle over negative detection angle at vertical plane，and the gas radiation on horizontal detection plane was greater than that on vertical detection plane.Under 90° vector state，the total radiation peak was only 43.3% of the non-vector state due to the shielding effect of deflected 3BSN structure on the front high-temperature components.The vector deflection of 3BSN reduced the gas velocity on the side far from curvature center and increased the temperature.A large local high temperature area appeared on the outside of the nozzle，which caused the wall radiation in negative detection angle range larger than that in positive detection angle on vertical detection plane，and the maximum relative difference was up to 71.9%.The deflection also obstructed the high temperature gas inside the nozzle.The negative detection angle of vertical detection plane can only cover the high temperature gas area at the nozzle exit，which was significantly smaller than the positive detection angle，resulting in the maximum relative difference of 20% in gas radiation.
Abstract: An internal penalty discontinuous Galerkin （IPDG） method was developed to solve the 2-D （two-dimensional）compressible Navier-Stokes equations together with chimera grid technique.The viscous flux was introduced as an auxiliary variable to reduce the equation order.The semi-discrete equation was solved with the discontinuous Galerkin method，and time marching was carried out with the implicit Newton-Krylov method.The accuracy of above methods was validated with the Couette flow in comparison of the exact solution.In addition，several test cases including viscous NACA0012 airfoil，steady and unsteady flow around a circular cylinder were simulated to verify the robustness and feasibility of the present methods.Eventually，the h-grid adaptive technique was also adopted in chimera grid to improve the resolution of shock wave.The inviscid transonic flow case over NACA0012 airfoil was also simulated.Result showed that，the number of cells after adaptation only increased by 8.4% compared with that without adaptation，but the shock resolution was significantly improved，and the calculated results near the shock were in better agreement with the experimental values，thereby verifying the effectiveness of the method.
Abstract: A method to improve the dynamic stress measurement accuracy for compressor blades was introduced.Dynamic stress calibration tests of the instrumented blades were carried out before assembly of the rotor.Strain gauge sensitivity ratios of the measured strain to the maximum modal strain for each vibration mode of the instrumented blades were acquired.The ratios were helpful in defining strain monitoring limits for the blades during engine tests，which made blade vibration safety monitoring more accurate.Advantages and drawbacks of two calibration methods，direct stress and af strength method （the product of vibration amplitude a multiplies vibration frequency f），were comparatively analyzed.Based on the direct stress method，stress calibration tests for two compressor blades were carried out.Results showed that the measured gauge sensitivity ratios of the 1# blade for the 1st mode and the 3rd mode were larger than the theoretical ratios with values of 17.4% and 24% respectively.And the gauge sensitivity ratio differences of the two blades （1# blade and 2# blade） for the 1st vibration mode reached 14.8%.It proved that calibration tests could improve measured data validation for blade dynamic stress comprehensively.
Abstract: In order to improve the fatigue resistance of the TC17 titanium alloy，shot peening on the surface was carried out.The effects of different shot peening intensities on fatigue resistance of TC17 titanium alloy were studied by means of rotary bending fatigue test，fracture analysis，residual stress field analysis，microstructure observation and surface integrity analysis before and after shot peening.The results showed that surface roughness of titanium alloy was increased from 0.315 μm to 0.5-1.0 μm after shot peening.Severe plastic deformation occurred on the surface during shot peening and the surface grains were refined.Residual compressive stress was introduced into the surface，the layer depth of residual compressive stress was about 125-151 μm，and the maximum compressive stress appeared about 30 μm in depth from the surface.The fatigue life of the rotary bending sample after 0.3 shot peening was better than that of other samples.Under the stress levels of 740 MPa and 840 MPa，the fatigue life of 0.3 shot peened sample increased by 4.5 times and 7.5 times than that of non-shot peened sample，respectively.The increase of fatigue life was attributed to the inhibition of fatigue crack initiation and propagation by the refined grains，high-density dislocations and residual compressive stress on the surface induced by shot peening.Excessive shot peening intensity caused the high surface roughness，delamination and microcrack formation on the surface，which resulted in the decrease of fatigue life of 0.35 and 0.4 shot peened samples.
Abstract: Using a combination of finite element analysis and computational fluid dynamics，a model for solving the leakage and flow characteristics of the brush seal considering the wear of the brush bristle was established，and an experimental device for the friction and wear characteristics of the brush seal based on the cylindrical circumferential friction was designed and built.The positive pressure characteristics between the brush bristle and the rotor were studied by experiment.Based on the experimental verification of the accuracy of the numerical simulation，the influence of the brush bristle wear on the leakage and flow characteristics of the brush seal was studied.The research results showed that from the upstream to the downstream of bristle pack，the positive pressure between the brush bristle and the rotor，and the wear length of the brush bristle both decreased first and then increased；the wear length of the brush bristle increased with the increase of the amount of interference and the diameter of the brush bristle，and decreased with the increase of pressure ratio and the length of the brush bristle.The leakage of the brush seal considering the wear of the brush bristle increased with the increase of the pressure ratio，the amount of interference and the diameter of the brush bristle，and decreased with the increase of the bristle length.The viewpoint was proposed that the change rate of leakage characterized the degree of influence of the brush bristle wear on the leakage characteristics of the brush seal.The change rate of the leakage increased with the increase of the diameter of the brush bristle，and decreased with the increase of the length of the brush bristle and the pressure ratio.Under pressure ratio of 2.5 conditions，when the diameter of the brush bristle increased from 0.09 mm to 0.11 mm，the change rate of leakage increased from 9.53% to 19.18%；when the length of the brush bristle increased from 15.556 mm to 17.556 mm，the change rate of leakage decreased from 13.47% to 5.38%.Increasing the length of the brush bristle and reducing the diameter of the brush bristle can reduce the influence of bristle wear on the sealing performance of the brush seal.
Abstract: In order to study the containment of double-layer metal-structured casing of aero-engines under blade`s impact，the ballistic test of double-layer titanium alloy laminated target plates with different clearances was carried out by using a sliding-bore gun test system.The process of blade penetration into target plate was recorded by high-speed camera system，and the damaged targets were recovered.The penetration resistance and failure modes of different laminated target plates were studied.Through 28 effective ballistic tests，it was found that： compared with the test results of the combination with the same thickness of the inner （attack surface） and outer target plate，the larger gap indicated the worse anti-penetration ability of the target plate；compared with the test results of the combination without the gap，the anti-penetration ability of the combination with the thinner inner target plate was better than that of the combination with the same thickness of the inner and outer target plates.The ballistic test was simulated numerically by commercial finite element software ANSYS/LS-DYNA.The finite element simulation results were in good agreement with the test results.It was found that the anti-penetration ability of the combination with thinner inner target was stronger than that of the combination with thicker inner target，and both of them were stronger than that of the combination with the same thickness of inner and outer targets；in the case of no clearance，the ballistic limit increased with the increase of blade angle of attack，but the angle of attack of the sudden increase of velocity curve was different.The dimensionless target thickness determined the failure mode of the blade impacting the double-layer targets.
Abstract: To solve the problem of insufficient prediction accuracy of the compressor performance by the original model in the through-flow analysis program，and to improve the reliability of the compressor through-flow analysis process，a compressor cascade performance database was established based on the numerical simulation results of a large number of multiple circular arc cascades.Based on this database，neural network modeling method was used to establish the baseline loss coefficient and baseline deviation angle models of compressor cascade.Results showed that，the prediction accuracy of the two models for the baseline loss coefficient and baseline deviation angle of cascade met the requirements of engineering applications，with the accuracy of ±0.002 and ±1°，respectively.During the verification process，it could be found that the neural network models significantly improved the prediction accuracy of both compressor's overall performance and the flow details，especially at the core flow region.Moreover，the improvement of the accuracy of baseline loss coefficient and baseline deviation angle had a positive effect on the prediction accuracy of loss coefficient and deviation angle at off-design conditions.
Abstract: Taking the multi-stage rotor and stator assembly process as the research object，and based on the spatial geometric transformation theory and the shape and position tolerance theory，a multi-stage rotor blade tip clearance prediction method was proposed.Specifically，the rotor and stator system was divided into three subsystems：rotor，stator and support.The sub-models of assembly deviation were established for each subsystem.The coordinates of each submodel were unified.The measured deviation data of fitting surface were taken as the input to calculate the rotor blade tip coordinates and casing coordinates at the same axial position，so as to calculate the rotor blade tip clearance.The test results showed that the prediction model could accurately predict the rotor blade tip clearance，and the maximum relative error of prediction was 11%，thus providing a reference for the assembly quality analysis.
Abstract: To numerically analyse the process of air-assisted atomization for diesel spray，the computational fluid dynamics simulation tool Fluent was applied.Three-dimensional fluid calculation model considering diesel injection process in fuel/air premixing chamber was established without changing the injector structure.The air-assisted injection process of diesel/air two-phase flow was investigated in constant volume chamber.The effects of air injection pressure，ambient backpressure and fuel temperature on diesel spray characteristics were investigated.The results showed that the air injection pressure had a great influence on the spray penetration distance.As the air injection pressure increased from 0.65 MPa to 0.75 MPa，the penetration distance increased by 12.2%，while the Sauter mean diameter （SMD） of diesel only decreased by 5.2%.When maintaining the SMD of fuel spray at an optimal level，the pressure ratio of 7.5 was beneficial to reduce the air consumption.Ambient backpressure had significant influence on penetration distance and SMD of fuel spray，as the pressure ratio was 1.875，the flow field at injector outlet almost did not produce supersonic flow，and the decrease of aerodynamic force significantly affected the breaking process of fuel droplets，making the SMD increase to 40.7 μm at the injection time of 4 ms.Fuel temperature had little effect on spray penetration distance，while it had a great effect on SMD.The SMD of fuel decreased significantly with the increase of temperature，and the SMD corresponding to 323 K decreased by 12.3% compared with that of 293 K.At the injection time of 4 ms，the overall SMD of fuel was 14.9 μm and the evaporated mass fraction of fuel reached 36.93%.
Abstract: This parameter was modified by calculating the entrainment flow rate.The result showed that the modified turbulence parameter had great control effect on flow state and total pressure coefficient.Both the total pressure coefficient and the effective swirl fraction relied on the modified turbulence parameter and the inlet pre-swirl fraction.In addition，the ability of air to maintain radial internal flow state was enhanced with the increase of modified turbulence parameter.Therefore，the area requiring suppress circumferential rotation decreased，which shortened the optimal length of the tubed vortex reducer with the minimum total pressure coefficient.
Abstract: In view of the speed control of the working point switching process of the extended-range auxiliary power unit （APU），a predictive control strategy based on the Hammerstein nonlinear model was proposed.The excitation data were identified by sparse least squares support vector machine-adaptive chaotic particle swarm optimization （SLSSVM-ACPSO） algorithm，and the Hammerstein nonlinear model of engine was established.When solving the optimal control sequence in model predictive control，the relaxation factor was used to relax the constraint boundary，and the active set method （ASM）-ACPSO combination algorithm was used to obtain the solution.The variable predictive time domain strategy was applied in the control process.The system simulation model was established，and the simulation results showed that in the processes of switching from the warming-up point to low load point and from the low load point to the medium load point，the stabilization times were 2.57 s and 2.77 s respectively，and the speed overshoot rates were 2% and 1.6% respectively，which were better than the two comparison strategies.In the process of switching from medium load point to the high load point，the speed overshoot rate was relatively larger，but the change of torque was more smooth in the control process.The simulation result showed that the model predictive control strategy for APU system had fast speed response，small speed and torque overshoot rate，and exhibited good dynamic control effect.
Abstract: To systematically analyze the dynamic characteristics of rocket sled travelling along rail，the three-dimensional Eluer-Bernouli beam elements were used to discretize the rocket sled system，and the nonlinear contact force model of the slipper-rail considering rail irregularity and slipper wear was established by reconstructing the long-range irregularity rail，and finally，the nonlinear dynamics equations was solved by Newmark-β combined with Newton-Raphson local iteration method to obtain the numerical solution of dynamic characteristics on rocket sled. Though experimental verification，the results showed that the peak value of the vertical overload to the mass of rocket sled was about 2 700g at Mach number of 2 speed；the duty cycle of rocket sled on rail was proportional to forward speed，and the vertical duty cycle of rocket sled was 16% at Mach number of 2 speed；the wear of slippers in high speed section accounted for 84% of the total trajectory wear，and the single contact wear of slipper-rail was proportional to the third power of the forward speed.
Abstract: In order to analyze the frequency characteristics of a LOX/LH2 rocket engine，an engine linearized dynamic model with main and subsidiary systems coupled was developed，including pipelines described by one-dimensional distributed parameter model，combustion components described by adiabatic flow model considering entropy wave and pumps components were described by model considering cavitation.The transcendental functions in thermodynamic component models and distributed parameter models were expressed by rational approximation.After establishing modular simulation model in Simulink，a frequency sweep was conducted to obtain the frequency characteristics，and hot test data was implemented in validation.The result showed that the modular simulation library is user-friendly，and is capable of analyzing the low to medium frequency characteristics of the engine system.Under the excitation of gas pulsation，the engine produced the frequency response of 10 and 165 Hz in main system and 124 and 204 Hz in subsidiary system.The calculated characteristics frequency showed better accuracy while considering the rotation speed feedback.The stability of the engine system was improved while increasing the local resistance of the pipeline before the pump or increasing the pressure drop of the gas generator injection.
Founded in 1986,
monthly
Sponsor: Aviation Industry Corporation of China
Sponsored by: Chinese Society of Aeronautics Beijing University of Aeronautics and Astronautics
