AR-based deep learning for real-time inspection of cable brackets in aircraft

In the process of aircraft assembly, there exist numerous and ubiquitous cable brackets that shall be installed on frames and subsequently need to be manually verified with CAD models. Such a task is usually performed by special operators, hence is time-consuming, labor-intensive, and error-prone. In order to save the inspection time and increase the reliability of results, many researchers attempt to develop intelligent inspection systems using robotic, AR, or AI technologies. However, there is no comprehensive method to achieve enough portability, intelligence, efficiency, and accuracy while providing intuitive task assistance for inspectors in real time. In this paper, a combined AR+AI system is introduced to assist brackets inspection in a more intelligent yet efficient manner. Especially, AR-based Mask R-CNN is proposed by skillfully integrating markerless AR into deep learning-based instance segmentation to generate more accurate and fewer region proposals, and thus alleviates the computation load of the deep learning program. Based on this, brackets segmentation can be performed robustly and efficiently on mobile devices such as smartphones or tablets. By using the proposed system, CAD model checking can be automatically performed between the segmented physical brackets and the corresponding virtual brackets rendered by AR in real time. Furthermore, the inspection results can be directly projected on the corresponding physical brackets for the convenience of maintenance. To verify the feasibility of the proposed method, experiments are carried out on a full-scale mock-up of C919 aircraft main landing gear cabin. The experimental results indicate that the inspection accuracy is up to 97.1%. Finally, the system has been deployed in the real C919 aircraft final-assembly workshop. The preliminary evaluation reveals that the proposed real-time AR-assisted intelligent inspection approach is effective and promising for large-scale industrial applications.     

Differential geometry modeling and application of roller pose and trajectory of robot roller hemming for complex curved surface-curved edge panels

Roller hemming is a relatively new process used to achieve high-precision assembly of auto-body enclosure panels. During the process of roller hemming, accuracy of the roller pose and trajectory affects the hemming quality of the product. The traditional passive method based on robot teaching to determine the pose of the roller is inefficient and time-consuming. In these studies, we proposed an active method for solving roller pose and trajectory based on differential geometry for curved surface-curved edge geometric characteristics of auto-body enclosure panels and multi-pass reciprocating motions of the roller. Firstly, the local coordinate system of the die was constructed based on the Frenet Frame according to the normal vector of the surface of die and the tangent vector of the curved die edge. Secondly, the coordinate system of the die, diameter of the roller, TCP-RTP value, and inclination of the roller were combined to form the roller pose based on a homogeneous transformation matrix. Based on the obtained trajectory curve of the roller reference point, the equal chord deviation error method was used to analyze the roller trajectory. Finally, a roller pose and trajectory solving algorithm was developed based and implemented using PYTHON to obtain the positions and poses of the roller at several discrete reference points. ABAQUS software was subsequently utilized to complete modeling of the roller pose and trajectory. This research supports the multi-field mechanical simulation of robot roller hemming for curved surface-curved edge panels and provides support for determining roller pose and kinematic trajectory of industrial robot roller hemming for curved surface-curved edge panels.     

Elastic-plastic-brittle transitions of potassium dihydrogen phosphate crystals: characterization by nanoindentation

Potassium dihydrogen phosphate (KDP) crystals are widely used in laser ignition facilities as optical switching and frequency conversion components. These crystals are soft, brittle, and sensitive to external conditions (e.g., humidity, temperature, and applied stress). Hence, conventional characterization methods, such as transmission electron microscopy, cannot be used to study the mechanisms of material deformation. Nevertheless, understanding the mechanism of plastic-brittle transition in KDP crystals is important to prevent the fracture damage during the machining process. This study explores the plastic deformation and brittle fracture mechanisms of KDP crystals through nanoindentation experiments and theoretical calculations. The results show that dislocation nucleation and propagation are the main mechanisms of plastic deformation in KDP crystals, and dislocation pileup leads to brittle fracture during nanoindentation. Nanoindentation experiments using various indenters indicate that the external stress fields influence the plastic deformation of KDP crystals, and plastic deformation and brittle fracture are related to the material’s anisotropy. However, the effect of loading rate on the KDP crystal deformation is practically negligible. The results of this research provide important information on reducing machining-induced damage and further improving the optical performance of KDP crystal components.The full text can be downloaded at https://link.springer.com/article/10.1007/s40436-020-00320-3     

仪表着陆系统的航道调整仿真系统研究

仪表着陆系统(ILS)为飞机安全着陆提供重要的保障。 针对现如今日益递增的航班数量,如何确保 ILS 高效、安全的运 行已成为民航部门面临的重要问题。 通过对调制深度差(DDM)的理论分析、研究,得出了影响 DDM 的主要因素是调制度因子 m90 ,在此基础上,采用控制变量法和最小二乘法分别建立了航道和调制度因子和载波负值的数学模型。 基于 NM7000 设备搭 建了可动态模拟航道调整过程的仿真系统。 最后将所建立系统与实际情况进行对比分析,验证所建立模型的准确性,数据表 明,仿真相对误差小于 1%,说明所提模型精度较高,可以为空中管制人员的参考依据提供理论支持。     

Multi-objective robust design of helical milling hole quality on AISI H13 hardened steel by normalized normal constraint coupled with robust parameter design

Helical milling is a hole-making process which has been applied in hardened materials. Due to the difficulties on achieving high-quality boreholes in these materials, the influence of noise factors, and multi-quality performance outcomes, this work aims the multi-objective robust design of hole quality on AISI H13 hardened steel. Experiments were carried out through a central composite design considering process and noise factors. The process factors were the axial and tangential feed per tooth of the helix, and the cutting velocity. The noise factors considered were the tool overhang length, the material hardness and the borehole height of measurement. Response models were obtained through response surface methodology for roughness and roundness outcomes. The models presented good explanation of data variability and good prediction capability. Mean and variance models were derived through robust parameter design for all responses. Similarity analysis through cluster analysis was performed, and average surface roughness and total roundness were selected to multi-objective optimization. Mean square error optimization was performed to achieve bias and variance minimization. Multi-objective optimization through normalized normal constraint was performed to achieve a robust Pareto set for the hole quality outcomes. The normalized normal constraint optimization results outperformed the results of other methods in terms of evenness of the Pareto solutions and number of Pareto optimal solutions. The most compromise solution was selected considering the lowest Euclidian distance to the utopia point in the normalized space. Individual and moving range control charts were used to confirm the robustness achievement with regard to noise factors in the most compromise Pareto optimal solution. The methodology applied for robust modelling and optimization of helical milling of AISI H13 hardened steel was confirmed and may be applied to other manufacturing processes.     

One-axis hysteresis motor driven magnetically suspended reaction sphere

This paper presents the design, modeling, control, and experimental results for a one-axis magnetically suspended reaction sphere (1D-MSRS) driven by a hysteresis motor. The goal of this work is twofold: (a) to conduct a preliminary study for magnetically suspended reaction sphere for three-axis spacecraft attitude control, and (b) study the potential of hysteresis motors for the reaction wheel/sphere drives. The 1D-MSRS uses a hysteresis motor with a spherical rotor made of solid steel. The rotor sphere is magnetically suspended in all translational directions, and is driven about the vertical axis by a bearingless hysteresis motor. We present the modeling and control of the magnetic suspension of the bearingless motor in the 1D-MSRS, and the hysteresis motor dynamics are analyzed by a hysteresis motor equivalent circuit model. The 1D-MSRS system has experimentally demonstrated a starting torque of 8.9 mNm under 0.7 A peak sinusoidal excitation current. With this excitation the sphere can run up to 12,000 rpm synchronously in the presence of air drag. This study demonstrates that the hysteresis motor has strong potential for use in high-speed, low-vibration reaction wheels.     

Numerical investigation on the failure phenomena of stiffened composite panels in post-buckling regime with discrete damages

The aim of this work is to investigate the final failure response of damaged composite stiffened panels in post buckling regime under compressive load, by using progressive failure analysis (PFA) methodology. The selected panel is characterized by T shaped stringers and it is representative of the upper skin panel, toward the wing tip, of the wing box of a typical regional aircraft. PFA methodology has been applied in order to predict in addition to the initiation of the local failure also its propagation up to the final collapse of the panel, in presence of local damage (barely visible impact damage, BVID) and in post-buckling regime. For this purpose, discrete damages have been considered in the skin of the panel. According to the indications contained in many guidelines finalized to the preliminary design of composite structures, a simplified design model of BVID has been considered in this work, in particular a hole 1/4 in. in diameter has been used to simulate this damage. The collapse load of the panel has been evaluated considering different locations of a single damage and also considering multi-damage maps (the latter are more representative of a real damage scenario). The results of PFA presented in this work illustrate the combined effect of the reduction of the panel stiffness and of the damage propagation, and the sensitivity of the buckling onset and of the residual strength of the panel with respect to different damage locations and damage density.     

Morphology and electrical properties of graphene–epoxy nanocomposites obtained by different solvent assisted processing methods

We report a comparative study of graphene nanoplatelets (GNPs)–epoxy nanocomposites with enhanced electrical conductive properties obtained with two different processing techniques. In the first one (TEC1), the epoxy monomer was added to a previously produced GNP–chloroform suspension and after the evaporation of the solvent, the hardener was added. In the second technique (TEC2), the hardener was added to a GNP–tetrahydrofuran suspension and after the evaporation of the solvent, the epoxy monomer was added. Although there was good dispersion of GNPs in the epoxy matrix with both techniques, the nanocomposite based on TEC1 showed a slightly better dispersion than the one based on TEC2. Electric and dielectric characterization showed that it is possible to reach the electrical percolation threshold at reasonably low GNP contents.     

Bionic digital brain realizing the digital twin-cutting process

A digital twin (DT) is an effective means of achieving physical and information fusion and has great potential for achieving a new paradigm of cutting process (CP) intelligence. This paper traces the relevant studies of digital technology in manufacturing and proposes a bionic digital brain (BDB) as the intelligent core of a digital twin-cutting process (DTCP) framework. The BDB was built with digital neurons (DN) as the basic functional unit, and the reaction mechanism between the DN stimulated the BDB to compute intelligently in real time (RT). The left brain obtains the prophetic theoretical processing information through the DN. The right brain receives actual perceptive processing information through the DN. After the corpus callosum fuses the left and right brain information, the DN outputs the optimal control solutions in RT to ensure demand. The details of monitoring, predicting, optimizing, and controlling the CP based on this framework are described in detail through a case study, demonstrating the powerful information processing capability of the BDB and the RT precision intelligent machining effects. The developed DTCP system confirmed the feasibility and effectiveness of the proposed framework. It provided a theoretical and technical reference for applying and promoting DT technology in the CP.