Transactions of Nanjing University of Aeronautics & Astronautics
Volume 0,Issue 2,2024 Table of Contents
2024(2):135-146.
DOI: 10.16356/j.1005-1120.2024.02.001
Abstract:
Residual stress （RS） within titanium alloy structural components is the primary factor contributing to machining deformation. It comprises initial residual stress （IRS） and machined surface residual stress （MSRS）， resulting from the interplay between IRS and high-level machining-induced residual stress （MIRS）. Machining deformation of components poses a significant challenge in the aerospace industry， and accurately assessing RS is crucial for precise prediction and control. However， current RS prediction methods struggle to account for various uncertainties in the component manufacturing process， leading to limited prediction accuracy. Furthermore， existing measurement methods can only gauge local RS in samples， which proves inefficient and unreliable for measuring RS fields in large components. Addressing these challenges， this paper introduces a method for simultaneously estimating IRS and MSRS within titanium alloy aircraft components using a Bayesian framework. This approach treats IRS and MSRS as unobservable fields modeled by Gaussian processes. It leverages observable deformation force data to estimate IRS and MSRS while incorporating prior correlations between MSRS fields. In this context， the prior correlation between MSRS fields is represented as a latent Gaussian process with a shared covariance function. The proposed method offers an effective means of estimating the RS field using deformation force data from a probabilistic perspective. It serves as a dependable foundation for optimizing subsequent deformation control strategies.
Residual stress （RS） within titanium alloy structural components is the primary factor contributing to machining deformation. It comprises initial residual stress （IRS） and machined surface residual stress （MSRS）， resulting from the interplay between IRS and high-level machining-induced residual stress （MIRS）. Machining deformation of components poses a significant challenge in the aerospace industry， and accurately assessing RS is crucial for precise prediction and control. However， current RS prediction methods struggle to account for various uncertainties in the component manufacturing process， leading to limited prediction accuracy. Furthermore， existing measurement methods can only gauge local RS in samples， which proves inefficient and unreliable for measuring RS fields in large components. Addressing these challenges， this paper introduces a method for simultaneously estimating IRS and MSRS within titanium alloy aircraft components using a Bayesian framework. This approach treats IRS and MSRS as unobservable fields modeled by Gaussian processes. It leverages observable deformation force data to estimate IRS and MSRS while incorporating prior correlations between MSRS fields. In this context， the prior correlation between MSRS fields is represented as a latent Gaussian process with a shared covariance function. The proposed method offers an effective means of estimating the RS field using deformation force data from a probabilistic perspective. It serves as a dependable foundation for optimizing subsequent deformation control strategies.
2024(2):147-157.
DOI: 10.16356/j.1005-1120.2024.02.002
Abstract:
The lack of key materials has emerged as one of crucial factors affecting the execution of helicopter assembly production plans. Accurate material delivery time prediction can guide assembly production planning and reduce frequent changes caused by material shortages. A lifelong learning-based model for predicting delivery time of materials is proposed on the basis of internal data sharing within the helicopter factory. During real-time prediction， the model can store new memories quickly and not forget old ones， which is constructed by gated recurrent unit （GRU） network layer， ReLU activation layer， and fully connected layers. To prevent significant precision degradation in real-time prediction， a regularization parameter constraint method is proposed to adjust model parameters. By using this method， the root mean square error （RMSE） in the model’s prediction on the target domain data is reduced from 0.032 9 to 0.013 4. The accuracy and applicability of the model for real-time prediction in helicopter assembly is validated by comparing it with methods such as L2 regularization and EWC regularization， using 25 material orders.
The lack of key materials has emerged as one of crucial factors affecting the execution of helicopter assembly production plans. Accurate material delivery time prediction can guide assembly production planning and reduce frequent changes caused by material shortages. A lifelong learning-based model for predicting delivery time of materials is proposed on the basis of internal data sharing within the helicopter factory. During real-time prediction， the model can store new memories quickly and not forget old ones， which is constructed by gated recurrent unit （GRU） network layer， ReLU activation layer， and fully connected layers. To prevent significant precision degradation in real-time prediction， a regularization parameter constraint method is proposed to adjust model parameters. By using this method， the root mean square error （RMSE） in the model’s prediction on the target domain data is reduced from 0.032 9 to 0.013 4. The accuracy and applicability of the model for real-time prediction in helicopter assembly is validated by comparing it with methods such as L2 regularization and EWC regularization， using 25 material orders.
2024(2):158-173.
DOI: 10.16356/j.1005-1120.2024.02.003
Abstract:
The thin-walled blade is a crucial component of aero engines， which is highly susceptible to significant deformation during the machining process. Existing research on deformation control focuses on reducing cutting force and machining-induced residual stress （MIRS）. The initial residual stress （IRS） generated in the process of heat treatment and forging is used to reduce the deformation of thin-walled parts under the influence of cutting force and MIRS. Because the IRS measurement is difficult and destructive， this paper proposes a reverse identification method of IRS to measure the IRS of Ti6Al4V. The proposed method is more consistent with the trend of stress and deformation distribution compared with the conventional method. To investigate and decouple the interplay between IRS， MIRS and cutting force on machining deformation， this study employs a curved blade for experimental validation and develops a finite element model to predict the deformation. It is found that cutting force accounts for 46.17% of the deformation with an average value of 26.36 μm， while MIRS accounts for 53.83% with an average value of 30.70 μm. Coupling IRS reduces MIRS maximum deflection deformation from 35.32 μm to 15.50 μm， which provides a new approach to optimize machining deformation through IRS distribution.
The thin-walled blade is a crucial component of aero engines， which is highly susceptible to significant deformation during the machining process. Existing research on deformation control focuses on reducing cutting force and machining-induced residual stress （MIRS）. The initial residual stress （IRS） generated in the process of heat treatment and forging is used to reduce the deformation of thin-walled parts under the influence of cutting force and MIRS. Because the IRS measurement is difficult and destructive， this paper proposes a reverse identification method of IRS to measure the IRS of Ti6Al4V. The proposed method is more consistent with the trend of stress and deformation distribution compared with the conventional method. To investigate and decouple the interplay between IRS， MIRS and cutting force on machining deformation， this study employs a curved blade for experimental validation and develops a finite element model to predict the deformation. It is found that cutting force accounts for 46.17% of the deformation with an average value of 26.36 μm， while MIRS accounts for 53.83% with an average value of 30.70 μm. Coupling IRS reduces MIRS maximum deflection deformation from 35.32 μm to 15.50 μm， which provides a new approach to optimize machining deformation through IRS distribution.
2024(2):174-183.
DOI: 10.16356/j.1005-1120.2024.02.004
Abstract:
The selection of heat source model is very important to accurately predict the distribution of temperature field and melting pool geometry in the numerical modeling of additive manufacturing process. The surface model， volumetric model and double-ellipsoid model are selected for comparison and analysis. These three heat source models are progarmmed as user-defined subroutines with Abaqus/Standard simulation software to predict the peak temperature and melting pool geometry during selective laser melting （SLM） of IN625. The comparison between simulation and experimental results shows that double-ellipsoid model can predict the melting pool geometry well， while the volumetric model provides comparative peak temperature predictions. In contrast， the surface model exhibits significant deviations in both melting pool geometry and peak temperature. The findings in this research highlight the need for model calibration or modification to enhance efficiency and accuracy before further research can be conducted.
The selection of heat source model is very important to accurately predict the distribution of temperature field and melting pool geometry in the numerical modeling of additive manufacturing process. The surface model， volumetric model and double-ellipsoid model are selected for comparison and analysis. These three heat source models are progarmmed as user-defined subroutines with Abaqus/Standard simulation software to predict the peak temperature and melting pool geometry during selective laser melting （SLM） of IN625. The comparison between simulation and experimental results shows that double-ellipsoid model can predict the melting pool geometry well， while the volumetric model provides comparative peak temperature predictions. In contrast， the surface model exhibits significant deviations in both melting pool geometry and peak temperature. The findings in this research highlight the need for model calibration or modification to enhance efficiency and accuracy before further research can be conducted.
2024(2):184-201.
DOI: 10.16356/j.1005-1120.2024.02.005
Abstract:
Wide-speed range flight is a critical design objective and development direction for hypersonic vehicles. However， the complex environmental changes pose design conflicts for fixed-configuration vehicles under different flight conditions. Hypersonic morphing vehicles can adapt to various flight conditions and meet performance requirements by presenting different configurations. This paper introduces numerical simulations to investigate the aerodynamic characteristics of a foldable-wing vehicle， and focuses on the lift-to-drag ratio， longitudinal static stability， and directional static stability of the aerodynamic configuration in different wing folding states at various flight altitudes and Mach numbers. The impact of varying wing folding angles （0°， 45°， 90°） on the aerodynamic performance are compared. The results indicate that across the entire range of speeds studied （Ma=0—5）， the smaller wing folding angles result in the higher lift coefficients， drag coefficients， and lift-to-drag ratios. The wing folding angle of 0° exhibits the highest lift-to-drag ratio. In terms of longitudinal stability， the configuration with a smaller folding angle has better longitudinal stability. As the Mach number increases， the differences in longitudinal stability between different folding angles initially decrease and then increase. The static stability margins change from 1∶0.95∶0.84 to 1∶0.98∶0.88， then to 1∶0.89∶0.79. In addition， configurations with larger wing folding angles exhibit better directional stability. All three wing folding configurations are directionally stable during the low-speed flight phase. As the Mach number increases， the 0° and 45° folding angles gradually become directionally unstable.
Wide-speed range flight is a critical design objective and development direction for hypersonic vehicles. However， the complex environmental changes pose design conflicts for fixed-configuration vehicles under different flight conditions. Hypersonic morphing vehicles can adapt to various flight conditions and meet performance requirements by presenting different configurations. This paper introduces numerical simulations to investigate the aerodynamic characteristics of a foldable-wing vehicle， and focuses on the lift-to-drag ratio， longitudinal static stability， and directional static stability of the aerodynamic configuration in different wing folding states at various flight altitudes and Mach numbers. The impact of varying wing folding angles （0°， 45°， 90°） on the aerodynamic performance are compared. The results indicate that across the entire range of speeds studied （Ma=0—5）， the smaller wing folding angles result in the higher lift coefficients， drag coefficients， and lift-to-drag ratios. The wing folding angle of 0° exhibits the highest lift-to-drag ratio. In terms of longitudinal stability， the configuration with a smaller folding angle has better longitudinal stability. As the Mach number increases， the differences in longitudinal stability between different folding angles initially decrease and then increase. The static stability margins change from 1∶0.95∶0.84 to 1∶0.98∶0.88， then to 1∶0.89∶0.79. In addition， configurations with larger wing folding angles exhibit better directional stability. All three wing folding configurations are directionally stable during the low-speed flight phase. As the Mach number increases， the 0° and 45° folding angles gradually become directionally unstable.
2024(2):202-217.
DOI: 10.16356/j.1005-1120.2024.02.006
Abstract:
The investigation of aircraft icing problems has led to a growing interest in mixed-phase icing. This paper used drag coefficient， adhesion， and icing thermodynamic models to calculate ice accretion on a quasi-three-dimensional multi-element airfoil under mixed-phase conditions. Firstly， mesh generation and airflow field calculations are conducted on the multi-element airfoil. Next， numerical simulations are carried out to analyze the characteristics of ice crystals impingement， adhesion， and ice accretion under mixed phase conditions. The results show that the adhesion mass flow rate of ice crystals is high and can pose a threat to flight safety， and adhesion may also occur in the runback water zone. Moreover， with the increase of liquid water content （LWC） over total water content （TWC）， ice crystals are more likely to adhere to surfaces and participate in surface icing.
The investigation of aircraft icing problems has led to a growing interest in mixed-phase icing. This paper used drag coefficient， adhesion， and icing thermodynamic models to calculate ice accretion on a quasi-three-dimensional multi-element airfoil under mixed-phase conditions. Firstly， mesh generation and airflow field calculations are conducted on the multi-element airfoil. Next， numerical simulations are carried out to analyze the characteristics of ice crystals impingement， adhesion， and ice accretion under mixed phase conditions. The results show that the adhesion mass flow rate of ice crystals is high and can pose a threat to flight safety， and adhesion may also occur in the runback water zone. Moreover， with the increase of liquid water content （LWC） over total water content （TWC）， ice crystals are more likely to adhere to surfaces and participate in surface icing.
2024(2):218-232.
DOI: 10.16356/j.1005-1120.2024.02.007
Abstract:
Logistics unmanned aerial vehicles（UAVs） have brought new opportunities for the expansion of the global express logistics industry， especially to effectively overcome the shortcomings of ground transportation. However， since logistics UAVs are still in their infancy， it is necessary to analyze the collision risk during their operation. Using the theory of collision modeling in conflict zones， this study examines the potential safety hazards of logistics UAVs flying in specific airspace according to their characteristics and limitations. First， to measure the impact of various factors such as reliability and failure rates on the safe operation of logistics UAVs in certain airspace， a collision risk analysis model between logistics UAVs and other drones in a specific airspace is established. Second， by analyzing the factors that affect the safe operation of logistics UAVs， including airspace conditions， human-machine systems， environmental conditions， and management conditions， a collision risk analysis model between logistics UAVs and civil aircraft operating in particular airspace is established. To verify the accuracy of the proposed models， the models in both cases are solved and compared with the safety target criteria established by the International Civil Aviation Organization （ICAO）.
Logistics unmanned aerial vehicles（UAVs） have brought new opportunities for the expansion of the global express logistics industry， especially to effectively overcome the shortcomings of ground transportation. However， since logistics UAVs are still in their infancy， it is necessary to analyze the collision risk during their operation. Using the theory of collision modeling in conflict zones， this study examines the potential safety hazards of logistics UAVs flying in specific airspace according to their characteristics and limitations. First， to measure the impact of various factors such as reliability and failure rates on the safe operation of logistics UAVs in certain airspace， a collision risk analysis model between logistics UAVs and other drones in a specific airspace is established. Second， by analyzing the factors that affect the safe operation of logistics UAVs， including airspace conditions， human-machine systems， environmental conditions， and management conditions， a collision risk analysis model between logistics UAVs and civil aircraft operating in particular airspace is established. To verify the accuracy of the proposed models， the models in both cases are solved and compared with the safety target criteria established by the International Civil Aviation Organization （ICAO）.
2024(2):233-243.
DOI: 10.16356/j.1005-1120.2024.02.008
Abstract:
Lift-type aircraft has the characteristics of wide flight airspace， large range of speed changes， strong maneuverability requirements， which makes it a research focus in recent years. Aiming at difficulties of large uncertainty and strong external disturbance in the control model of lift-type aircraft， an improved discrete-time sliding mode variable structure control method based on disturbance observer is presented. Firstly， the discrete control model of the pitch channel is established. Secondly， an improved discrete-time sliding mode variable structure control system based on disturbance estimation is designed and its stability is proved. Thirdly， the influence of parameter selection on control system is analyzed， and the validity of control law is verified. Finally， the comparison with the traditional discrete sliding mode variable structure control system is carried out. The control system’s simulation results demonstrate that， compared with the traditional discrete-time sliding mode control system， the proposed control system has higher control accuracy， stronger robustness， faster convergence to zero and less chattering. The proposed control law can effectively improve the flight quality， and can realize stable control for complex mission of lift-type aircraft.
Lift-type aircraft has the characteristics of wide flight airspace， large range of speed changes， strong maneuverability requirements， which makes it a research focus in recent years. Aiming at difficulties of large uncertainty and strong external disturbance in the control model of lift-type aircraft， an improved discrete-time sliding mode variable structure control method based on disturbance observer is presented. Firstly， the discrete control model of the pitch channel is established. Secondly， an improved discrete-time sliding mode variable structure control system based on disturbance estimation is designed and its stability is proved. Thirdly， the influence of parameter selection on control system is analyzed， and the validity of control law is verified. Finally， the comparison with the traditional discrete sliding mode variable structure control system is carried out. The control system’s simulation results demonstrate that， compared with the traditional discrete-time sliding mode control system， the proposed control system has higher control accuracy， stronger robustness， faster convergence to zero and less chattering. The proposed control law can effectively improve the flight quality， and can realize stable control for complex mission of lift-type aircraft.
2024(2):244-252.
DOI: 10.16356/j.1005-1120.2024.02.009
Abstract:
This paper studies the physical layer security performance of non-orthogonal-multiple-access （NOMA） communication system. When the base station incorporates the down-link NOMA scheme to send information， due to the openness of the channel， the information is easy to be eavesdropped， and when there are multiple randomly distributed eavesdroppers， the security performance will be further reduced. In order to enhance physical layer security performance of the system with hardware impairments， we take the method named Protected Zone into account. Based on the characteristics of the direct link between ground users and base station， we apply the Rician fading to model small-scale fading. We also assume the locations of multiple eavesdroppers follow homogeneous Poisson point process （HPPP）. The closed-form expressions of average secrecy capacity are derived with the help of Gaussian-Chebyshev quadrature， and the asymptotic expressions are also provided for obtaining the insight under high signal-to-noise-ratio cases. The simulation results verify the effectiveness of the protected zone method for enhancing the security performance， and also illustrate the impact of different parameters on the secrecy performance of the system.
This paper studies the physical layer security performance of non-orthogonal-multiple-access （NOMA） communication system. When the base station incorporates the down-link NOMA scheme to send information， due to the openness of the channel， the information is easy to be eavesdropped， and when there are multiple randomly distributed eavesdroppers， the security performance will be further reduced. In order to enhance physical layer security performance of the system with hardware impairments， we take the method named Protected Zone into account. Based on the characteristics of the direct link between ground users and base station， we apply the Rician fading to model small-scale fading. We also assume the locations of multiple eavesdroppers follow homogeneous Poisson point process （HPPP）. The closed-form expressions of average secrecy capacity are derived with the help of Gaussian-Chebyshev quadrature， and the asymptotic expressions are also provided for obtaining the insight under high signal-to-noise-ratio cases. The simulation results verify the effectiveness of the protected zone method for enhancing the security performance， and also illustrate the impact of different parameters on the secrecy performance of the system.
2024(2):253-262.
DOI: 10.16356/j.1005-1120.2024.02.010
Abstract:
With the rapid development of space technology， the situation awareness ability of spacecraft is increased. As compared to the optical sensors， inverse synthetic aperture radars （ISARs） have the capability of high-resolution imaging in all day from far range regardless of the light condition. Furthermore， the component recognition is much desired by the accurate evaluation of the threat degree of surrounding spacecrafts. In this paper， we propose a multitask-you only look once （Multitask-YOLO） network based on the YOLOv5 structure for recognition and segmentation of solar panels of satellite ISAR images. Firstly， we add a segmentation decoupling head to introduce the function of segmentation. Then， the original structure is replaced with spatial pyramid pooling fast （SPPF） to avoid image distortion， and with distance intersection over union （DIoU） to speed up convergence. The accuracy of segmentation and recognition is improved by introducing an attention mechanism in the channels. We perform the experiments using simulated satellite ISAR images. The results show that the proposed Multitask-YOLO network achieves efficient and accurate component recognition and segmentation. As compared to typical recognition and segmentation networks， the proposed network exhibits an approximate 5% improvement in mean average precision （mAP） and mean intersection over union （mIoU）. Moreover， it operates at a higher speed of 16.4 GFLOP， surpassing the performance of traditional multitask networks.
With the rapid development of space technology， the situation awareness ability of spacecraft is increased. As compared to the optical sensors， inverse synthetic aperture radars （ISARs） have the capability of high-resolution imaging in all day from far range regardless of the light condition. Furthermore， the component recognition is much desired by the accurate evaluation of the threat degree of surrounding spacecrafts. In this paper， we propose a multitask-you only look once （Multitask-YOLO） network based on the YOLOv5 structure for recognition and segmentation of solar panels of satellite ISAR images. Firstly， we add a segmentation decoupling head to introduce the function of segmentation. Then， the original structure is replaced with spatial pyramid pooling fast （SPPF） to avoid image distortion， and with distance intersection over union （DIoU） to speed up convergence. The accuracy of segmentation and recognition is improved by introducing an attention mechanism in the channels. We perform the experiments using simulated satellite ISAR images. The results show that the proposed Multitask-YOLO network achieves efficient and accurate component recognition and segmentation. As compared to typical recognition and segmentation networks， the proposed network exhibits an approximate 5% improvement in mean average precision （mAP） and mean intersection over union （mIoU）. Moreover， it operates at a higher speed of 16.4 GFLOP， surpassing the performance of traditional multitask networks.
Mobile website
Website Copyright © Transactions of Nanjing University of Aeronautics & Astronautics