2016 Vol. 31, No. 4

Display Method:
Calibration technology of high-temperature thermocouple for combustor exit of an aero-engine
YANG Can, WU Wei-li, XIONG Yi-bin, YAO Zheng-rong
2016, 31(4): 769-774. doi: 10.13224/j.cnki.jasp.2016.04.001
Abstract:
The high-temperature thermocouple for measuring aero-engine combustor exit temperature were calibrated in the high-temperature wind tunnel, and radiation correction coefficients were experimentally determined for two types of non-cooled and air-cooled thermocouples. Results of the calibration experiment presented that the radiation errors decreased with increasing airflow velocity and pressure, and it became greater when the temperature difference increased between the airflow temperature and aero-engine case wall temperature. Comparing the two types of thermocouple, the measurement errors of the non-cooled high-temperature thermocouple were smaller than the air-cooled high-temperature thermocouple.
Experimental investigation of temperature measurement on the wall of flame tube for a triple combustor
XIONG Qing-rong, ZHU Guo-cheng, ZHONG Ming
2016, 31(4): 775-779. doi: 10.13224/j.cnki.jasp.2016.04.002
Abstract:
In order to study the best cooling style for a triple combustor of three different cooling styles, temperature-sensitive paints and thermocouple were used to study of wall temperature profiles of the flame tube. The result show that, it is difficult to distinguish cooling effect when using a few thermocouples to measure the wall temperature; the best cooling style of flame tube could be analyzed by temperature-sensitive paints, including wall temperature and temperature distribution; the temperature recognition accuracy based on isotherm temperature identification is ±10℃, the experiment result is an important reference to compare and assess the cooling effect of flame tube.
Enhanced heat transfer characteristics in a large aspect ratio rectangular channel with inclined perforated rib
ZHU Qiang-hua, CUI Miao, GAO Xiao-wei
2016, 31(4): 780-787. doi: 10.13224/j.cnki.jasp.2016.04.003
Abstract:
The enhanced heat transfer characteristics of inclined perforated rib in a large aspect ratio rectangular passage were studied through numerical simulations. The holes inclination angle and entrance flow Reynolds number were adjusted in the range of 0°-30° and 3×104-9×104 respectively. The variation laws of channel friction factor and its rib-wall Nusselt number with these parameters were system analyzed. The calculation results reveal that the inclined perforated rib has a lower friction factor and an improved local heat transfer immediately downstream from the perforated rib due to the impingement of jet-like flows as compared with the solid rib. With the increasing of holes inclination angle, the relative friction factor is elevated monotonously, and the averaged Nusselt number ratio take on up-down tendency, therefore an optimal holes inclination angle is 15°, when the thermal performance factor reached its maximum. The friction factor slightly decreases and the averaged Nusselt number gradually increases with the increasing of Reynolds number for the inclined perforated rib.
Charge-effect on the homo-heterogeneous condensing flow
CUI Ke, LIU Hua-ping, SONG Yan-ping, CHEN Huan-long, CHEN Fu
2016, 31(4): 788-799. doi: 10.13224/j.cnki.jasp.2016.04.004
Abstract:
Based on the Fletcher's nucleation theory and the two-fluid model for homo-heterogeneous condensation, the effect of charge-effect on the condensing flow was discussed. The results show that the formula of the nucleation rate calculation with or without charge-effect both have high resolution. The calculated results are in good agreement with the experimented results and the relative error is not than 5%, indicating that the two-fluid model for homo-heterogeneous condensation has high precision. When the particle is not charged, the difference of supercooling in the flow field before and after the input of particles with 5nm radius is small. Although the input of 8nm and 10nm radius particles reduces the peak value of the supercooling as 2K and 7K respectively, it also causes the downstream migration of spontaneous condensation. When the particles are charged with a unit of electrical charge, i.e. Q=1e, the input of 5nm radius particles shows most obvious effect on decreasing the nucleation energy barrier and improving the heterogeneous nucleation rate. This effect becomes weaker and even weakest with the input of 8nm and 10nm radius particles respectively. As the charge rise up to three units of electrical charge, the input of 5nm and 8nm radius particles reduce the peak value of supercooling as 10K and 6K compared with homogeneous condensation respectively, which also obviously inhibit the occurrence of homogeneous condensation.
Penetration characteristics and optimization of transverse jet on condition of vortex generator
ZHAO Yan-hui, LIANG Jian-han, YAN Tian-tian
2016, 31(4): 800-806. doi: 10.13224/j.cnki.jasp.2016.04.005
Abstract:
Vortex generator (VG) was placed at the upstream of jet orifice. Nano-particle plane laser scattering system(NPLS) and particle image velocimetry(PIV) technology were employed to observe the flow field of transverse jet on condition of VG with five different configurations. Penetration depth of transverse jet was extracted and compared. Penetration depth was enhanced 30%-60% higher by placing VG at the upstream of jet orifice. Sensitivity analysis of VG configuration parameters, jet to mainstream momentum ratio and distance between VG and jet orifice was proceeded. It turns out that jet to mainstream momentum ratio plays a dominate role in jet penetration depth. Those parameters were optimized based on target function. It is revealed that the top value of dimensionless jet penetration depth is 12.68 while VG half cone angle is 36°, VG gradient angle is 6°, dimensionless distance between VG trailing edge and jet orifice is 38.8 and jet to mainstream momentum ratio reaches 10.6.
Numerical study on the effect of parallel injection on film-cooling effectiveness at turbine blade tip
JIANG Wei, XIE Dan-mei
2016, 31(4): 807-815. doi: 10.13224/j.cnki.jasp.2016.04.006
Abstract:
For the purpose of improving the film-cooling effectiveness on blade tip region of gas turbine, numerical simulation was carried out on the blade tip model to investigate the flow and heat transfer for four kinds of cooling holes. It was found that the covered structure can double the film-cooling effectiveness compared to the uncovered structure for aclinic injection structure.The physics was revealed by analysis of the flow filed, and results showed that the aclinic injection can improve flow distribution, obtain better film-coverage and more uniform flow distribution, consequently increasing the film-cooling effectiveness.The effect of gap width, cover degree, cover thickness and fence on film-cooling effectiveness was studied, and the results showed that the increase of the cover degree and decrease of the gap width can get a higher film-cooling effectiveness, but the cover degree and gap width were restricted by structural strength, technological level and secondary flow inlet pressure.Aclinic cover can obtain a better film-cooling effectiveness than the inclined one and there is a best cover thickness enabling to obtain the highest film-cooling effectiveness for the aclinic one.Sufficiently long fence can improve the flow distribution and increase the film-cooling effectiveness.
Influence of total temperature and pressure at the inlet of convergent exhaust system of turbofan engine on temperature distribution at plume centerline
WANG Feng, JI Hong-hu, YU Ming-fei
2016, 31(4): 816-822. doi: 10.13224/j.cnki.jasp.2016.04.007
Abstract:
During the overall process of designing the turbofan engine, the model for predicting the signature of infrared radiation was established.With the method for calculating the fluid dynamics, the influence of the total temperature and total pressure at the inlet of convergent exhaust system of turbofan engine on the temperature distribution at plume centerline was estimated; the overall process of designing the turbofan engine was required and the approximate formula for calculating the temperature distribution at plume centerline was achieved. The result shows that the dimensionless distance corresponding to the dimensionless temperature at plume centerline is proportional to the total pressure of the inner duct, the bypass duct and total temperature of the bypass duct at the inlet of convergent exhaust system of turbofan engine, but inversely proportional to the total temperature of the inner duct.Therefore, it can be concluded that the summed-up experience formula of temperature distribution at the plume centerline of turbofan engine agrees well with the numerical results.
Characteristics of the recirculation zone downstream a dual-swirl cup at varied inlet Reynolds numbers
YU Bo-wen, ZHANG Chi, YANG Qian, XU Quan-hong, LIN Yu-zhen
2016, 31(4): 823-829. doi: 10.13224/j.cnki.jasp.2016.04.008
Abstract:
To analyze the role of the flow structure downstream the swirl cup on the performance of aero gas turbine combustors, PIV (particle image velocimetry) was used to measure the non-reacting flow field downstream a dual-swirl cup under ambient pressure and temperature, and the effect of the inlet Reynolds number on the recirculation zone downstream a dual-swirl cup was studied. The experimental results indicate that the flow structure downstream is a converge-diverge type when the inlet Reynolds number is low. As the inlet Reynolds number increases to 25000, the flow structure turns to a diverge type.Then the flow structure of recirculation zone remains unchanged with increasing inlet Reynolds number, but the normalized velocities in the recirculation zone would increase accordingly.
Effects of film hole arrangement on central-cone cooling and infrared radiation characteristics of exhaust system
WANG Xu, ZHANG Jing-zhou, SHAN Yong
2016, 31(4): 830-835. doi: 10.13224/j.cnki.jasp.2016.04.009
Abstract:
The axisymmetric convergent-expansion nozzle exhaust system of a certain turbofan engine was taken as a prototype wherein the film cooling schemes on the central-cone surface were numerically investigated by CFD/IR methods. The cooling schemes were compared and the cooling action of central-cone surface on the infrared radiation characteristics was revealed. The results show that the film cooling on the central-cone surface could effectively decrease the surface temperature, and suppress the infrared radiation intensity between 0 degree and 15 degree viewed from backward direction. The best cooling scheme achieved about 282K decrease on the central-cone surface as well as 57% relative decrease of infrared radiation intensity on 0 degree direction by comparing with the prototype exhaust system.
Boundary layer control of ultra-high-lift low pressure turbine blade with surface roughness
SUN Shuang, LEI Zhi-jun, LU Xin-gen, LI Wei, FU Yu, DONG Li-hui
2016, 31(4): 836-846. doi: 10.13224/j.cnki.jasp.2016.04.010
Abstract:
The effects of surface roughness on the aerodynamic performance and boundary layer condition of low pressure turbine blade PACKD-A were investigated experimentally under steady state and unsteady state. The upstream wakes were simulated by the rotating bars, thus, an unsteady environment state was formed. To produce the roughness blades, the sandpaper was stuck into the slot, which was incised on the smooth blade. The profile loss and the loading of the suction side were measured by aerodynamic probes and static pressure holes associated with the pressure transducer. And the details of the boundary layer were measured with a single hotwire probe. Result shows that the roughness strips covering 5.2% suction surface length (starting at 44.3% suction surface length, stopping at 49.5% suction surface length) and with 8.82μm roughness height have a better performance than others in the unsteady state. The profile loss is reduced with this roughness strip with the whole tested Reynolds number(3×104~12×104). In addition, the roughness strips covering 19.5% suction surface length (starting at 30% suction surface length, stopping at 49.5% suction surface length) and with 20.91μm roughness height reduce the profile loss and extended the operating margin of the turbine blade with low Reynolds number(less than 8×104) under steady state. However, some additional profile loss is produced with this type of roughness with high Reynolds number(more than 8×104).
Experiment on variable tip clearance characteristics of compressor cascade with suction-side winglet
HAN Shao-bing, ZHONG Jing-jun
2016, 31(4): 847-856. doi: 10.13224/j.cnki.jasp.2016.04.011
Abstract:
In order to further explore the effect mechanisms of suction-side winglet under different tip clearance sizes, an experimental investigation on the variable tip clearance characteristics of the compressor cascade with/without suction-side winglet was carried out. Results showed that at 1% relative tip clearance size, the tip leakage flow suppressed three-dimensional separation on the corner of suction surface and endwall, the cascade loss and endwall blockage were both minimum; and this clearance size was found to be optimum tip clearance. The suction-side winglet reduced the strength of tip leakage vortex but resulted in a stronger passage vortex, and the three-dimensional separation zone was formed again. For the particular tip clearance, the loss and endwall blockage of cascade with winglet were both higher than the baseline tip cascade. In case of 2% and 3% relative tip clearance sizes, for the cascade with suction-side winglet, the tip separation became stronger, and the strength of tip leakage vortex was reduced. In case of these two tip clearance sizes, the total pressure loss coefficients of the cascade was reduced by 8.9% and 12.5%, respectively by the suction-side winglet, and the blockage coefficients with winglet were 6.9% and 6.3% lower compared with the cascade without winglet. For these three tip clearance sizes, the compressor cascade performance sensitivity to tip clearance variation was improved by the suction-side winglet, and the degree of outlet flow angle's underturning and overturning induced by leakage vortex was reduced by the suction-side winglet.
Effect of blade sweep on inlet flow in transonic compressor rotors
CHANG Hao, ZHU Fang, JIN Dong-hai, GUI Xing-min
2016, 31(4): 857-866. doi: 10.13224/j.cnki.jasp.2016.04.012
Abstract:
NASA Rotor 37 was selected as a baseline blade. Sweep features at tip regions were formed by moving the airfoil sections above the 90% span along the chordwise direction. There were five cases including the baseline rotor, two forward rotor and backward sweep rotors in total. A numerical simulation softwave was used to simulate the five rotors and obtain the overall characteristics and three-dimensional (3-D) flow fields. The results indicated that the tip forward sweep resulted in a decrease of inlet averaged incidence, while the backward sweep caused the reverse consequence. A circumferentially averaged method was introduced to provide the circumferential fluctuation terms as well as values in the inviscid momentum equation. The distributions of the circumferential fluctuation terms in the radial inviscid momentum equation at the blade inlet were calculated for comparison between the five cases, and were analysed for incidence variation. Forward sweep at tip regions resulted in the increase of axial velocity and reduction of circumferential velocity, further reducing the incidence due to the change of flow equilibrium at blade inlet, while backward sweep caused an inverse effect in the same region. The effect of blade sweep on flow fields at blade inlet was studied from the quasi-3-D perspective.
Mechanism for the effect of lean on the inlet flow of a single blade row
XU Han, ZHU Fang, JIN Dong-hai, GUI Xing-min
2016, 31(4): 867-876. doi: 10.13224/j.cnki.jasp.2016.04.013
Abstract:
The effect of lean on the inlet flow of a single blade row was studied with the dimensionality reduction method, and the mechanism for lean improving endwall condition was explained by the inlet flow change. The circumferential fluctuation(CF) source items were obtained through the derivation of the circumferentially averaged Navier-Stokes(N-S)equations. The radial and circumferential components of the CF sources existed yet before the blade passage with a uniform incoming flow, representing the magnitude of the inlet CF. Different leaned cascade models were numerically calculated and circumferentially averaged. The results indicate that lean could change the magnitude and distribution of the CF source items before the blade passage, leading to the radial migration of the inlet flow. The positive lean diverts the inlet flow from hub to tip and improves the endwall condition, while the negative lean has a contrary effect.The inlet CF sources are also affected by the aerodynamic parameters such as the inlet Mach number, blade solidity and stagger angle. Neither an over-high inlet Mach number nor an over-wide stagger angle is good for the effect of blade lean on the inlet flow.
Occurrence mechanism of partial surge
PAN Tian-yu, SUN Tai-lu, LI Zhi-ping, LI Qiu-shi
2016, 31(4): 877-885. doi: 10.13224/j.cnki.jasp.2016.04.014
Abstract:
To fully understand the instability mechanism of highly load compressors, the occurrence mechanism of instability inception partial surge was studied in detail. Based on the experimental data of instability evolution at different rotation speeds and uniform intake of a transonic compressor, the instability evolution under inlet distortion was studied by experimental and numerical methods. The results show that under uniform intake, partial surge can only be detected at high rotation speed while at low rotation speed partial surge can't be detected. So it can be supposed that high blade loading at hub is related to the occurrence of partial surge. The results under different inlet condition suggest that lower blade loading in the hub region could prevent the occurrence of partial surge at high rotation speed while partial surge would occur at the low rotation speed if the blade loading in the hub region is higher. High blade loading in the hub is proved as a necessary condition for the occurrence of partial surge. In this compressor, if the diffusion factor exceeds 0.6 at near stall point, partial surge may occur.
Influence of wakes on boundary layer stability in low-pressure turbines
LIANG Yun, LIU Huo-xing, ZOU Zheng-ping
2016, 31(4): 886-893. doi: 10.13224/j.cnki.jasp.2016.04.015
Abstract:
As the most important disturbance source, the wakes dominate the instability and transition process of boundary layer in low-pressure turbines. A detailed experimental study on the velocity fluctuation in a laminar separated shear layer on the flat plate of low-pressure turbine pressure distribution with the hot-wire probe at Reynolds number of 130000 were presented. The disturbance development under stable and instable incoming conditions was compared to analyze the influence of the wakes on the destabilization mechanism. Results show that the disturbance grows linearly in the former part of the separation bubble, and nonlinear interaction appears when the disturbance saturates, leading to vortex shedding at the maxima height of the bubble. Inviscid K-H(Kelvin-Helmholtz) mechanism dominates the instability process in the separation bubble. The occurrence of convective linear instability mechanisms was also revealed when amplifying the velocity fluctuations carried by wakes while propagating downstream. This accelerated the transition process and suppressed the separation bubble.
Numerical simulation on noise reduction for the cascade with trailing edge serrations
TONG Fan, QIAO Wei-yang, JI Liang, WANG Liang-feng, XU Kun-bo, WANG Xun-nian
2016, 31(4): 894-902. doi: 10.13224/j.cnki.jasp.2016.04.016
Abstract:
Large eddy simulation and acoustic analogy were performed to investigate the effect of trailing edge serrations on the turbine cascade noise. Two different trailing edge were designed. Comparisons were made for the sound power of two different serrated trailing edge cascades and straight trailing edge cascade under Re=3.3×105 (based on the blade chord and cascade outlet velocity).The result shows that the trailing edge serrations can reduce the blade suction surface boundary layer's separation noise by about 5dB and the trailing edge vortex shedding noise by about 10dB. Further investigation shows that trailing edge can reduce blade surface pressure fluctuation amplitude by about 50% near the trailing edge and change the vortexes of large spanwise correlation length into those of smaller spanwise correlation length. Furthermore, the trailing edge serrations can eliminate the trailing edge vortex shedding. The above three factors work jointly and contribute to the final noise reduction.
Design and test of a jet flap for a large amphibian
SUN Wei-ping, YANG Kang-zhi, QIN He-jun
2016, 31(4): 903-909. doi: 10.13224/j.cnki.jasp.2016.04.017
Abstract:
According to the requirements of a large amphibian's operating features and performance based on the original wing, a high-lift device design technology of boundary layer control jet flap was studied.The scheme of boundary layer control jet flap was designed. By using computational fluid dynamics numerical simulation as a means to assess the preliminary design, the design scheme's aerodynamic characteristics and flow field structure were analysed and evaluated, and finally wind tunnel tests were conducted to verify the lift effect of the scheme.The result of the design scheme showed that in a wide range of blowing momentum coefficient, the maximum lift coefficient had varying degrees of growth; when the blowing momentum coefficient was 0.2, the incremental maximum lift coefficient was 1.0; according to the influence of the prototype slipstream projections, when the boundary layer control jet flap was employed, the take off and landing speed could drop about 30%, achieving the design target.
Numerical simulation for optimal design of wide angle diffuser in limited condition
CONG Cheng-hua, REN Ze-bin, YANG Gao-qiang, WANG Ning
2016, 31(4): 910-917. doi: 10.13224/j.cnki.jasp.2016.04.018
Abstract:
In order to improve the static pressure recovery of specific low speed wind tunnel wide angle diffuser(WAD) and reduce the total pressure loss to the level of the splitter plates, different configurations of WAD were studied using computational fluid dynamics (CFD). The screens could restrain flow separation. The position and opening area ratio of the screens had an important effect on the flow regimes and total pressure loss. The total pressure loss using linear wall WAD was greater than the expected value. Due to the oversize diffuser angle, the first screen played a little role on separation control of WAD inlet. By use of cubic curve wall, the total pressure loss evidently decreased and was less than the expected value. But a little separation in WAD existed. Simple segment linear walls had profound function on the drop of total pressure loss. Based on the method of maximum static pressure recovery, WAD using three segment linear walls was designed. The separation of inlet disappeared. And the total pressure loss was reduced to a great extent. Further investigation on the diffuser of three segment linear walls indicated that the angle of the first level diffuser was upper limit to the edge of separation based on the method of maximum static pressure recovery. To restrain separation of the third level diffuser, the method was reasonable to shorten the second level diffuser length and reduce the third level diffuser angle. Combination of satisfactory parameters, which could significantly produce less total pressure loss to the expected level, was obtained based on flow regime analysis of different configurations. It is concluded that the optimal design of WAD can be obtained based on numerical simulation.
Numerical investigation on influence of reverser flow to engine inlet flow field
WANG Zhi-qiang, SHEN Xi-gang, HU Jun, GAO Xiang, LIU Li-ping
2016, 31(4): 918-926. doi: 10.13224/j.cnki.jasp.2016.04.019
Abstract:
By solving the three-dimensional Reynolds averaged Navier-Stokes (N-S) equations, the detailed flow field information of the reverser flow was obtained under different taxiing-velocity conditions during the landing process of a large transport aircraft. The single engine calculation shows that in no case the reverser flow will be re-ingested by the engine. The aircraft/engine integrated calculation indicates that with the decrease of relative wind Mach number, the possibility of the reverser flow being re-ingested by engine keeps increasing. When the relative wind Mach number decreases to 0.1, the reverser flow is re-ingested by the outboard engine. Meanwhile, the inlet flow field of the engine is distored, and the circumferential steady total pressure distortion coefficient increases obviously. When the relative wind Mach number decreases to 0.05, the reverser flow is re-ingested by both the outboard and inboard engines. When the relative wind Mach number decreases to 0, the reverser flow is not re-ingested. But the reverser flow hinders the free flow and reduces the total pressure of the free flow ingested by the engines, finally causing the inlet flow field distortion of the engine.
Numerical solutions of adjoint Green's function for complex geometries
XU Xi-hai, LI Xiao-dong, HU Fang-qiang
2016, 31(4): 927-933. doi: 10.13224/j.cnki.jasp.2016.04.020
Abstract:
In order to simplify the grid generation during aerodynamic noise prediction, the computational aeroacoustic (CAA) method and immersed boundary method (IBM) were combined together to solve the adjoint Green's function. According to the basic formulation of adjoint Green's function, a special case of cylindrical acoustic scattering problem was designed. The comparison between the numerical results and the analytical solutions prove that the numerical methods used are reliable and feasible for the calculation of adjoint Green's function for problems with complex geometries. Finally, the combined numerical methods were used to solve the adjoint Green's function of the jet flow from a nozzle with chevron and pylon interaction. Result shows that, because of the scattering effect caused by jet flow shear layer, the distribution of adjoint Green's function in jet flow is uneven, and the maximum value is 3 times higher than the minimum value. This simulation results reveal the relative importance of different regions with the same sound intensity to the far field noise. Therefore, low noise nozzle design can focus on reducing the sound intensity in the larger area of Green's function.
Dynamic modeling of helicopter-flexible rope-slung load system based on Kane's method
WAN Shao-feng, CAO Yi-hua, HUANG Lei
2016, 31(4): 934-940. doi: 10.13224/j.cnki.jasp.2016.04.021
Abstract:
A dynamic model of a helicopter-flexible rope-slung load system based on Kane's method was completed. Through kinematical description of the system, the system was divided into three parts, i.e., helicopter ontology, slung rope and slung load.The helicopter ontology is a six-degree-of-freedom rigid body. The slung rope was simulated by a spring-connected lumped mass model; the slung load was simulated by a three-degree-of-freedom mass.Kane's dynamical equations of the helicopter-flexible rope-slung load system were established in the consideration of gravitational forces, aerodynamic forces and elastic forces.For the helicopter ontology and the helicopter-flexible rope-slung load system in consideration of flexible rope, the trim states at different forward velocities were computed first, then the motion modes were analyzed and compared.Simulation results show:the helicopter needs extra collective pitch with introduction of slung load; convergent/divergence rate of helicopter phugoid motion slows down in consideration of flexible rope; Kane's method is suited for building up the dynamic model of helicopter-flexible rope-slung load system, the equations are simple, shipshape and suited for programmable calculation.
Improved correction methods of aircraft engine fan speed based on similarity theory
XIA Cun-jiang, WANG Dan
2016, 31(4): 941-947. doi: 10.13224/j.cnki.jasp.2016.04.022
Abstract:
One improved correction method of aircraft engines fan speed by variable index factor was proposed based on the similarity theory. The cruise data of an aircraft engine from different airlines were collected to build the data sample. Method of support vector regression machine was used to establish the model of index factor and atmosphere temperature, and the model parameters were optimized based on genetic algorithm. The variable index factor model, consisting of indicated fan speed and atmosphere temperature, was set up to calculate the corrected fan speed. The model was applied by sample data, and the results were compared with those by fixed index factor method, then the error analysis was drawn through the calculation results of two methods. The result indicates that aircraft engine corrected fan speed has higher accuracy through the method of variable index factor and the model has good generalization performance, so it is an effective approach for computing aircraft engine corrected fan speed, providing a practical guidance for the aircraft engine performance prediction.
Aero-engine performance seeking control based on DMOM algorithm
WANG Yuan, LI Qiu-hong, HUANG Xiang-hua
2016, 31(4): 948-954. doi: 10.13224/j.cnki.jasp.2016.04.023
Abstract:
A distributed migration optimization method (DMOM) for multi-peak optimization problem was proposed to find global optima solution. In this algorithm, the local optima could be searched out by continuous discrete and migration with new individuals. Then the global optima solution was found by sorting. During the discrete process the possibility of sinking into local optima was reduced compared with other algorithms. The DMOM was used for aero-engine performance seeking control in digital simulation. Results show that, in minimum fuel mode and minimum turbine temperature mode, the DMOM is improved by more than 2 times compared with genetic algorithm (GA) and particle swarm optimization (PSO), the optimization accuracy of DMOM is improved by more than 60% compared with self-organizing migrating algorithm (SOMA), and more than 20% compared with feasible sequential quadratic programming(FSQP). Comparative results show that the calculation time can be reduced and the aero-engine performance can be improved by using this algorithm.
Estimation of unknown input observer disturbance matrix for fault diagnosis of turboshaft engine sensors
YANG Chao, WANG Xi, SHI Rui-jun, ZHOU Jian-bo, GAO Shi-huang, TAN Dao-liang
2016, 31(4): 955-964. doi: 10.13224/j.cnki.jasp.2016.04.024
Abstract:
Focusing on the research of unknown input observer of good performance to deal with the problem of control system's sensors fault, a method using a full-order state observer to estimate the disturbance matrix of turboshaft engine system model was proposed based on the analysis of relationship between the disturbance matrix reflecting modeling uncertainty and the accuracy of states observed by UIO. This method contributed to an accurate disturbance matrix so that the results of UIO had no combination of modeling uncertainty when the operating state of turboshaft was fixed. Furthermore, this method was developed to meet the requirements of variable operating states of the engine so that the modeling uncertainty had no influence on the design of UIO in this situation. Finally, some simulations were made to verify the proposed methods, which contribute to obtain a better matrix to ensure that the observed output residual of UIO is less than 2.
Turbo-shaft engine modeling method based on ADE-ELM
JIAO Yang, LI Qiu-hong, ZHU Zheng-chen, LIAO Guang-huang
2016, 31(4): 965-973. doi: 10.13224/j.cnki.jasp.2016.04.025
Abstract:
A modeling method based on adaptive differential evolution-extreme learning machine (ADE-ELM) was proposed to solve the balance equations for turbo-shaft engine models. The model achieved high accuracy and real time properties. The extreme learning machine (ELM) was trained by the data got from the Newton-Raphson (N-R) iteration model. The inputs were the residuals of balance equations before iteration, and the outputs were the adjustments of guess values after iteration. The parameters of the ELM were optimized by adaptive differential evolution (ADE) algorithm to enhance the mapping accuracy of the guess values adjustments. The sigmoid type adaptive scaling factor was adopted by ADE algorithm to improve its optimization ability. Taking the N-R iteration model as a criterion, the simulation results of ADE-ELM model under different flying states show that, the maximum modeling error is about 1/3 that of the single iteration per time step, and the time consumption is about 1/3 that of the single iteration per time step. The modeling method is valid.
Reliability evaluation method based on failure data with poor information
XIA Xin-tao, ZHU Wen-huan, MENG Yan-yan, QIU Ming
2016, 31(4): 974-985. doi: 10.13224/j.cnki.jasp.2016.04.026
Abstract:
Under the condition of failure data with poor information, the maximum entropy model for evaluation of mechanical product reliability was put forward. According to the empirical value formula of reliability, the reliability empirical value vector of failure data was obtained, and discrete failure frequency vector of lifetime data could be inferred by the empirical value vector (i.e. statistical histogram was obtained); based on the interval-mapping Newton iteration method, the maximum entropy probability density function was established, the failure probability distribution function could be gained by integration, and then the estimated true value function of reliability was acquired. Studies on simulation cases and experimental cases have proved the method proposed is not only able to evaluate the reliability with a known distribution but effective and feasible in the reliability assessment under the condition of failure data without any priori information about the possibility distribution. At the given lifetime, the difference between the values of reliability obtained via the maximum entropy method and the known distributions is very small, which only reaches 3.40%.
Impact of support stiffness on dynamic load sharing characteristics of planetary train system
ZHU Zeng-bao, JIANG Zhi-xiang, YIN Min
2016, 31(4): 986-992. doi: 10.13224/j.cnki.jasp.2016.04.027
Abstract:
The dynamic differential equation of the planetary train system was set up considering planetary carrier micro displacement, time-varying meshing stiffness, revolution damping and components gravity. The Modelica language for modeling, analyzing and optimizing the engineering system of various fields was used to solve the problem, and the impact of support stiffness of the main components on dynamic load sharing characteristics of this train was analyzed. The research results show that under the condition of rigid support, the load between the planet gears is seriously uneven at a smaller system error. When the support stiffness of one or more components is less than 107N/m, a good load sharing effect of this train could be achieved. The load sharing characteristic of this train is deteriorated rapidly with the increase of this component support stiffness within the support stiffness sensitive range of corresponding component. The load sharing effect for reducing the supporting stiffness of multiple components is better than that of single ones. The load is redistributed between planetary gears while the support stiffness of a planetary gear is changed. When the support stiffness of a planetary gear is cut down, its own load is reduced.
Reliability analysis of gear vibration based on dimensionalityreduction visualization and Kriging
YANG Li, TONG Cao
2016, 31(4): 993-999. doi: 10.13224/j.cnki.jasp.2016.04.028
Abstract:
To solve the problems of large computation and low precision during gear vibration reliability analysis, a reliability analysis method based on dimensionality reduction visualization and Kriging was proposed. Sample points were generated by Monte Carlo method. These points were transformed into two-dimensional pole feature space, and then Kriging model was used to predict the dividing line of safe and failure regions. When predicting the dividing line, an active learning approach of selecting points was introduced to establish Kriging model so that the utilization rate of sample points was improved dramatically, thanks to the properties of nonlinear prediction and error estimation of Kriging. Through gear vibration reliability analysis, and by comparing with traditional dimensionality reduction visualization technique, it is shown that the number of calls to the performance function changes from 975 numbers to 149 numbers, and calculation time changes from 12400s to 1810s. What's more, the result of this method is consistent with that of 100000 Monte Carlo simulation, so the efficiency and correctness is validated.
Flutter investigation of small aspect ratio compressor rotor blade
WU Chang-bo, ZHOU Bai-hao, LU Qing-fei, CUI Hai-tao
2016, 31(4): 1000-1005. doi: 10.13224/j.cnki.jasp.2016.04.029
Abstract:
For the characteristics of small aspect ratio compressor rotor blade, two judgement methods on the basis of energy method were presented. The analysis results of structure dynamics were employed to analyze the rotor frequency dispersion, vibration mode and resonance vibration, while the possibility of high mode coupled flutter was estimated. Then the vibration displacements from structure dynamics into aerodynamic mesh were applied to calculate three dimensional nonsteady viscous filed to get unsteady aerodynamic work, estimating the possibility of stall flutter. Finally, considering two analysis results as well as the test results and fracture analysis, cracks appeared in small aspect ratio compressor rotor blade within a short time due to the combined effects of stall flutter and coupled flutter.
Experiment on mechanical properties of resin matrix composites laminates under various temperatures
SONG Jian, WEN Wei-dong
2016, 31(4): 1006-1018. doi: 10.13224/j.cnki.jasp.2016.04.030
Abstract:
The mechanical properties of resin casting body and carbon fiber reinforced resin composite laminate were investigated under various temperatures experimentally. Influence of temperature on the stress-strain curve was discussed and then fracture surface analysis was also conducted. The experiment results showed that the tensile and compression properties of the resin casting body were obviously affected by temperature, and both of the properties were unequal. The average tensile modulus and strength fell by 31.73% and 44.71% compared with the value at room temperature (RT 20℃), and then dropped by 21.15% and 20.37%, respectively if continuing to heat up to 200℃. The average compressive modulus and strength at 160℃ decreased by 26.67% and 44.40% in comparison with those at RT, and the relative values continued to decline by 6.66% and 12.40%. For the composites, the longitudinal tensile property was hardly affected by temperature. The maximum amplitudes in the average stiffness and strength were only 2.82% and 2.53%. The longitudinal fractured surface morphology illustrated a "brushy" shape at RT, while a split configuration was observed at elevated temperature. The mechanical properties of transverse and in-plane shear were significantly affected by temperature and the stress-strain curve showed an obvious nonlinearity, however, a neat fractured surface subjecting to transverse loading and shear fractured surface had no significant necking appearance, which indicated brittle-natured fracture. The modulus of transverse and in-plane shear was reduced by 32.96%, 41.25% and the related strength declined by 15.83%, 30.96% at 160℃ compared with those at RT. When the temperature reached 200℃, the modulus continued to decline by 16.83%, 22.52%, and the corresponding strength dropped by 12.24% and 11.01%, respectively.
Self-pressurization modeling of nitrous oxide tank
SONG Chang-qing, XU Wan-wu, CHEN Jian
2016, 31(4): 1019-1024. doi: 10.13224/j.cnki.jasp.2016.04.031
Abstract:
Based on the three regions assumption of Zilliac, a self-pressurization model was established with the use of Peng-Robinson state of equation. An experimental apparatus of self-pressurization was constructed to verify the model. The result indicated that the saturation temperature was between the vapor and liquid temperatures, and more adjacent to the liquid temperature during the self-pressurization progress. The model showed a good performance in predicting the pressure and liquid temperature, but the formed temperature stratification in vapor region made it inaccurate to apply the lumped parameter method to model, so a model with more nodes in the vapor region was expected in future.