The hot-stamping quenching process is an advanced forming technique for thin-walled aluminum alloy parts with complex shapes, but is limited by the long standard T6 treatment time, such as no shorter than 12 h for 2219 aluminum alloy. In this paper, the fast two-step aging heat treatment process of 2219 aluminum alloy was investigated, by aging in dies with high-low temperatures to replace standard T6 treatment, and the effect of various aging conditions on the mechanical properties of the alloy was experimentally investigated. Firstly, the fast two-step aging experimental investigation was carried out to explore the effect of the first-step aging temperature and time, and the second-step aging time on the mechanical properties of the alloy. The results show that the yield and the ultimate strength of the alloy show a whole trend to decrease as the first-step aging temperature increases from 220 ℃ to 305 ℃ and the strength of the alloy increases as the second-step aging time increases from 0.5 h to 4 h. Then, according to the results of the fast two-step aging experiments, the range of the aging parameters was optimized, and based on this, an experimental study of the in-die aging process was carried out. The results show that the yield strength is not lower than 91% of the T62 state and the ultimate strength is not lower than 100% of the T62 state when the aging condition is 220 ℃/230 ℃/240 ℃×5 min+175 ℃×4 h. When the aging condition is 240 ℃×5 min + 175 ℃×4 h, the tensile strength of the alloy increases from 393 MPa to 419 MPa as the die contact pressure increases from 0.33 MPa to 1.82 MPa. Finally, the TEM analysis of several sets of specimens of the typical aging conditions including the aging condition of 240 ℃×5 min + 175 ℃×4 h was carried out to observe the morphology and distribution of the precipitations. The two-step high and low temperature ageing heat treatment process proposed in this paper reduces the total aging time by over 63% compared to the traditional aging process (175 ℃×12 h or 175 ℃×18 h) on the premise of ensuring the properties of the alloy, and significantly shortens the ageing cycle of thin-walled high-strength aluminum alloy stamped parts after forming.

In order to solve the problem of large sample demand and low computational efficiency in the current sensitivity analysis method, a global sensitivity analysis method based on polynomial chaos expansion was proposed. Firstly, a complete spatial error model was established based on the screw theory by taking the AC type double turntable five-axis CNC machine tool as the research object. Secondly, the polynomial chaos expansion model of machine tool geometric error was constructed. The orthogonal matching pursuit was used to sparse the model, and the Sobol sensitivity index based on this method was given. Furthermore, the geometric errors of five-axis CNC machine tools were analyzed. he approximate probability distribution of 41 errors are measured and counted, and the key geometric errors affecting he pose error components in each direction are analyzed. Compared with Monte Carlo simulation and Latin hypercube ampling, the correctness of the polynomial chaos expansion method is verified. Under the premise of not reducing the calculation accuracy, the sample size is reduced from 1×10⁵ to 1×10³, the calculation time is reduced by 96.8% and 98.1% respectively, and the calculation efficiency is significantly improved.

In order to make up for the shortcomings of the traditional inspection methods of rivet unevenness in precision, efficiency and stability, a new inspection method is proposed based on structured light and binocular vision. First, the structured light is projected onto the surface by an industrial projector, and then the  ojection image is captured by a binocular camera. Then the riveted area is extracted according to the edge, and the phase information of the area is calculated with the improved Gray code phase unwrapping. Feature points are matched based on phase information and point clouds are calculated according to the principle of disparity. Finally, the rivet unevenness is calculated by processing point data. The experimental results indicate that the absolute error is within ± 20 μm with repeatability less than 3 μm. The measurement time for a single rivet is approximately 2.2 s. Requirements for on-site measurement accuracy and efficiency can be met.

In order to address the shortcomings of low calibration accuracy and efficiency, as well as cumbersome manual operations, in the large-scale calibration of plug gauges in aviation manufacturing enterprises, a method for measuring the diameter of smooth plug gauges based on machine vision and laser caliper is proposed. The diameter of the gauge is measured by a laser caliper, and the machine vision module synchronously detects its posture in the light curtain, and corrects the measured values of the caliper. At the same time, the use of specialized tooling combined with an air flotation motion mechanism achieves automatic switching of multiple gauges and multi-position automatic measurement of a single plug gauge. The experimental results indicate that the repeated measurement standard deviation of the system is 0.16 μm. Absolute measurement error is less than 0.5 μm compared to the measured value of the vertical optical comparator. The measurement time for single end of the gauge is less than 45 s, which can meet the fast and high-precision verification requirements of commonly used plug gauges in production.

Aiming at the problem of strong reflection and multiple reflection in 3D measurement of industrial shiny parts, a technology based on high-dynamic range N-step fringe projection profilometry and single pixel imaging were developed, which effectively solved the problems of point cloud missing and accuracy degradation. Two methods were used to measure standard gauge blocks. The root mean square error of method based on high-dynamic range N-step fringe projection profilometry is 0.0066 mm. The root mean square error of method based on single pixel imaging is 0.0046 mm. With the step standard as the target, the absolute accuracy verification experiment was carried out. The absolute deviation is less than 0.007 mm. The proposed method was used to measure aero-engine impeller, turbine blade, and carbon fiber structures. Experimental results demonstrated that the proposed method combining high dynamic range N-step fringe projection profilometry and single pixel imaging can avoid point cloud missing. The complete data can be obtained in both strong reflection and multiple reflection areas. The proposed method has realized accurate 3D measurement of industrial shiny parts.

In order to automatically realize detection of aircraft skin damage, a machine vision detection method based on the improvement of channel redundancy for YOLOv7 is proposed. Firstly, aiming at the characteristics of the single background for the aircraft skin damage dataset, an improved algorithm of enhanced neck feature fusion is proposed, which improves the recognition accuracy and detection speed of aircraft skin damage. Secondly, in order to solve the problem of convolution channel redundancy for the backbone feature extraction network, PConv (Partial convolution) is introduced, and the lightweight of the backbone feature extraction network is proposed, which reduces the number of parameters for the model and improves the efficiency of damage identification. In the experimental part, different improved algorithms of enhanced neck feature fusion were first explored on the aircraft skin damage dataset, and the optimal improvement method was determined. Then, ablation and comparative experiments were carried out on the aircraft skin damage dataset, and compared with the original YOLOv7 algorithm, the mAP (Mean average precision) is increased by 2.3%, the FPS (Frames per second) is increased by 22.1 f/s, and the number of model parameters is decreased by 34.13%. Finally, the improved YOLOv7 model is compared with the mainstream object detection model, which proves the advanced nature of the improved algorithm.

The new generation of aircraft and other major aviation models presents new characteristics of high stealth, heavy load, long range, long life and other leaps and bounds, which makes the existing digital assembly simulation coordination technology based on theoretical model difficult to meet its ultra-high assembly accuracy and service performance requirements. In view of the new demand for further improvement of assembly accuracy, a virtual preassembly technology that introduces the measured model of key assembly characteristics is introduced. By accurately expressing the manufacturing error of parts and the accumulated error of assembly during the assembly process, the precise quality control and traceability adjustment of assembly key parts with high accuracy requirements are realized. By analyzing the characteristics of this technology, the framework of virtual pre-assembly technology system for aircraft based on measured data is constructed. The development status of key technologies related to the identification and analysis of key assembly characteristics of complex aircraft structures, the measurement and reconstruction of geometric features of parts oriented to virtual pre-assembly, and the virtual pre-assembly analysis oriented to assembly semantics and geometric feature constraints are summarized. Combined with the ultra-high dimensional accuracy and service performance requirements of the new generation of aircraft, the new development trend of virtual pre-assembly technology towards the actual model under the digital twin system is clarified.

In the five-axis flank milling process, it is difficult to calculate the instantaneous chip thickness due to the complex change of tool axis attitude. In order to improve the prediction accuracy of milling force, Firstly, the cutter tooth trajectory of cylindrical cutter is described by the geometric model of five axis flank milling. After establishing the micro element cutting force model, a calculation method of instantaneous cutting thickness is proposed, which is to calculate the mapping distance between the cutting point on the cutting edge and the rotary cylinder surface of the previous cutter teeth. Under the homogeneous coordinate transformation, the intersection calculation of line and plane in space is transformed into the intersection calculation of line and arc in two-dimensional plane. After considering the influence of tool runout on arc trajectory, the instantaneous chip thickness value is obtained by solving the equations. Finally, the milling force experiment was carried out on a five-axis machine tool. By comparing the measured force data, the simulation results are in good agreement with the measured values both in trend and size, which verifies the effectiveness of the established cutting force prediction model.

To vigorously promote the integration of information and industrialization, we should achieve technological breakthroughs in key fields of aerospace equipment. In the next decade, China's demand for aircraft will continue to increase, which requires aircraft parts to be manufactured with high efficiency and fast response time. But the research and application of the production line scheduling algorithm are still relatively few, which leads to the actual production line can not play the maximum efficiency. Aiming at the above problems, this paper takes the actual flexible production line as the research object, establishes the production scheduling model in line with the actual field production, and then uses the optimized particle swarm optimization algorithm to solve the model and gets the optimal solution.

In order to analyze the influence of the milling process parameters on the residual stress of the machined surface of aluminum alloy, according to the related theories of metal cutting finite element analysis, the aluminum alloy 6061 was used as the workpiece material, and the finite element model of milling was established. The single factor method is used to simulate the mirror milling of aluminum alloy 6061, and the influence of different spindle speed, milling depth and milling path on the surface residual stress of aluminum alloy 6061 is analyzed. The research shows that during the processing of aluminum alloy 6061, the influencing factors on the surface residual stress of the workpiece in descending order are milling depth < milling path < spindle speed. The spindle speed has the greatest influence on residual stress of the machined surface, and the optimal group of mirror milling reduces the deformation displacement by 33%  compared with ordinary milling.

The TC4 samples formed by selective laser melting (SLM) have poor fracture toughness, low cycle fatigue performance and obvious anisotropy. The TC4 samples were treated by the combination of cyclic annealing (at 700 – 950 ℃ respectively) and solid solution aging. The effects of heat treatment on the microstructure and mechanical properties of the TC4 samples were studied by optical microscope, scanning electron microscope and low cycle fatigue testing machine. The microstructure of SLM TC4 sample is composed of martensite α′ and martensite α″, the average value of fracture toughness is 36.4 MPa·m^0.5, and the anisotropy of fracture toughness is 25.7%. After heat treatment, α lath is separated from each other, equiaxed α phase and secondary α phase are formed. The fracture toughness of SLM TC4 is 96.0 MPa·m^0.5 after cyclic annealing at 700 – 950 ℃ and aging at 550 ℃, and the anisotropy of fracture toughness is 1.4%. Compared with the forged samples, the heat treated SLM sample exhibits better low cycle fatigue performance when the strain amplitude is greater than or equal to 0.9%; When the strain amplitude is less than 1%, the low cycle fatigue performance of the SLM sample produced by cyclic annealing at 920 ℃ and solid solution aging at 550 ℃ is higher than that produced by cyclic annealing at 920 ℃.

Aiming at the problem of untimely detection of end-effector chatter during robotic grinding of aeroengine blades, a method of chatter detection based on permutation entropy during grinding is proposed. Using the algorithm for sliding windows to calculate the continuous value of the permutation entropy of the original chatter signal, the empirical threshold of 0.95 for permutation entropy is used to determine whether chatter occurs at the end of the robot. Among them, the algorithm for the permutation entropy based on the ordinal pattern is adopted, which greatly improves the extraction efficiency of the permutation entropy. When the length of the signal is 10000, the computation time of the permutation entropy of the proposed algorithm is reduced to about 0.25 s. And the calculation time of proposed algorithm is reduced by an order of magnitude compared with the traditional algorithm of the permutation entropy. Numerical simulation and experimental results show that the proposed method can detect chatter 0.48 s earlier than the moment of chatter outbreak, which provides more reaction time for taking measures to suppress chatter.