Search

Obtaining clean, intact, and clear laser stripe images from weld laser stripe images with noise is a key issue in robot welding systems based on laser vision sensors, especially for robotic multi-layer multi-pass welding. However, existing methods typically require image pairs, consisting of images with noise and corresponding images without noise, which are very difficult to obtain. Meanwhile, methods that do not require image pairs suffer from low accuracy. Therefore, a model called DARM (denoising and restoration model) based on CycleGAN is proposed to denoise and restore laser stripe images for robotic multi-layer multi-pass welding in this paper. Firstly, a conditional generator assisted by a weld pass number is established to guide the generator in processing images by introducing a prior information. At the same time, the above generator is streamlined through a lightweight convolution module to reduce the model scale. Then, a double-ended discriminator is constructed to assess the authenticity and class of the processed images, enhancing the processing effect through additional class loss. Finally, experiments show that the RMSE, PSNR, and SSIMof DARM are 1.384, 25.457, and 0.977, respectively, and the average processing time is 9.0 ms, which can achieve accurate real-time denoising and restoration of multi-layer multi-pass weld laser stripe images. The proposed method does not require image pairs for learning, the processing effect is better than that of existing methods, and the model scale is significantly reduced.

The 9-degrees-of-freedom (DOF) machine is used to lay the rotating mold driven by the positioner. In order to reduce the 9-DOF machine to a 6-DOF non-redundant structure, this paper proposes a redundancy resolution method based on synthesized global and local performance indices. Since the number of non-redundant machine types is enormous, all non-redundant machines are divided into six classes depending on the redundant joints, called axis assignment schemes. After defining the global and the local performance indices of the non-redundant machine, the redundancy resolution method is proposed. It consists of three parts: Firstly, refine the machine types of each axis assignment scheme based on the average global indices. Then, select the best axis assignment scheme by comparing the scheme’s local indices for a given mold. Finally, optimize the non-redundant machine type contained in the optimal scheme with the objective of superior local indices and continuous positioner rotation space. In the experiment, using the air-inlet as an example, two motion planning methods gave different motion trajectories for the 7-DOF layup system. The effectiveness of the proposed method was verified analytically by comparing the joint motion and fiber placement speed of the optimal machine type with other types.

Automated fiber placement (AFP) is a key technology in the fabrication of composite aeronautical components, but the layup defects associated with uneven and insufficient contact pressure limit its application on irregular surfaces. Existing methods for pressure optimization in AFP are mostly based on extensive compaction experiments, which lacks theoretical basis and is not applicable to complex surfaces. In this paper, an analytical model of pressure distribution in AFP is proposed, which is suitable for any irregular surfaces. The influence of mechanical properties of contact materials and the mold curvatures on the calculation of contact pressure is taken into consideration. The Hertz theory is also extended to the three-dimensional contact area to suit the practical needs. Furthermore, an algorithm for pressure calculation along any layup path is proposed to rapid construct the pressure field. Then three sets of experiments are carried out on two aeronautical molds to validate the proposed model. The results show that the model can accurately calculate the contact pressure distribution in AFP, and compared with the previous model, the comprehensive prediction accuracy of the peak pressure is improved by 27.25 %. Combined with the layup path information, this model can be used to analyze the mechanical mechanism of layup defects, which provides a theoretical basis for the optimization of AFP parameters based on pressure distribution.

This paper presents a novel kinematic modeling and error parameter identification method for a six-axis gantry automated fiber placement (AFP) machine. Multi-body system theory is used to model the kinematics of the AFP machine, and then use the homogeneous transformation matrix to represent the transfer relationship of the coordinate system. Based on 54 geometric errors of the AFP machine, a volumetric error transfer model is established. To eliminate the training bias caused by sampling randomness, 10-fold cross-validation combined with the Levenberg–Marquardt method is used to train the kinematic model of the AFP machine. Based on the coordinates of random points in space measured by a laser tracker, all the position-dependent geometric errors and position-independent geometric errors of the linear and rotary axes can be identified simultaneously. After the error parameters are identified, the average error of the random points in the X-direction is reduced from 0.72 to 0.08 mm, a drop of 88.9%. Corresponding experiments are carried out on the body diagonals and the winglet placement path. The experimental results show that the volumetric error transfer model and error parameter identification method proposed in the article can accurately predict the volumetric errors of the AFP machine, and can also meet the requirements of the AFP machine placement accuracy.

Layup temperature is a vital factor that affects the production quality of automated fiber placement for thermoset composites. For high quality production, this paper proposes a new prediction model of layup temperature by considering the structure of infrared heater of automatic fiber placement equipment. The model was first verified by comparing with the temperature results of layup experiments. Then a prediction on the effects of heater structure arrangement on layup temperature was conducted in the model. The research shows that the length of infrared tube and the radian of reflective-film have significant influences, rather than other parameters. According to this investigation, a heater structure-based heating strategy was presented in the paper to efficiently achieve constant temperature control in variable speed layup process.

Automated fiber placement (AFP) combining with the autoclave curing process is the main manufacturing method for large-scale rotary composite parts. During the forming process, the mandrel plays an important role that affects the forming quality of the part directly. In this study, design and analysis of the mandrel structure for a composite S-shaped inlet are carried out. The preforming process of the rotary composite part using the AFP technique is introduced first. Then, design principles of the shaft are presented and applied to the shaft design for the mandrel of the S-shaped inlet, which then is transformed into an optimization problem. The internal penalty function method is adopted to solve this optimization problem in order to obtain the optimal axis position and axle diameter. Moreover, detailed structure design of the mandrel is proposed which takes the demolding into consideration, and the wall thickness is designed considering the total weight and the stiffness. The static analysis and modal analysis are also carried out by the finite element method to verify the feasibility of the proposed mandrel structure.

Automated fiber placement machine is the key equipment for low-cost and automated manufacturing of high-performance carbon composite materials. In order to meet the required position accuracy of fiber placement, this paper focuses on the kinematic modeling and parameter identification of the automated fiber placement machine. A kinematic model taking account of geometric deviations is established firstly. Since joint interfaces are the main origin of gravity deformation in a machine tool, an elastic beam deformation model is introduced to represent the joint interface, and then the former kinematic model is modified by analytical expressions of the gravity deformation for each joint interface. Based on the measurement data and the Levenberg-Marquardt optimization method, the parameter identification of the kinematic model is realized, and main issues such as measurement data selection, objective function definition are discussed. Finally, a kinematic calibration experiment is performed, and the experimental results verify the feasibility and validity of the modeling method. The position errors in Zdirection of the automated fiber placement machine are effectively reduced by over 80%, which suggests that the proposed method reduces the effect of the gravity deformation and improves the accuracy of the automated fiber placement machine.

To ensure the riveting quality and avoid riveting defects such as overdriven head or underdriven head, elastic and plastic mechanical analysis of the riveting process has been studied and a riveting energy control method is proposed in this paper. According to the target size of the driven head, the material properties, and the interference, the total riveting energy is calculated. Then, the relation of riveting duration and riveting energy is established to realize the control of the riveting impact energy in practical engineering. Furthermore, plastic dynamic impact buckling of the rivet is also analyzed to get impact buckling critical speed. Due to the dual-robot posture adjustment and countersunk self-centering ability, the working axes of the riveting tools, rivet, and hole are found to be well coaxial. The analysis shows there is little phenomenon of rivet bending in the dual-robot pneumatic riveting system and the impact buckling critical speed is much higher than the piston impact speed. Finally, finite element model is built for riveting dynamic simulation and dual-robot pneumatic riveting cell for fuselage panel assembly is developed to obtain experimental results. The comparison shows the proposed method has a good performance. The research results enhance the knowledge of energy effect of dual-robot pneumatic riveting, and thus offer positive guidance for practical application in aircraft assembly.

In order to investigate the effect of shim compensation for nonuniform gaps in aircraft assembly, the influence of the shims with different material and parameters on bolted joints is studied in this paper. According to the real material and assembly conditions of the aircraft joint structures, the specimen and experiment are designed to obtain the tensile performance of the joint structures with different shims. A three-dimensional finite element model, which incorporates the Johnson–Cook material property of the alloys, traction-separation law of liquid shims, contact relationships between the joint elements, and boundary conditions of the tensile process, is established with the specimen configurations. After validating through comparing with the experimental results, the modeling method is adopted to simulate the tensile response of the bolted joints with shims. Furthermore, both the influence of the shim material and thickness on the mechanical behaviors of bolted joints is investigated in detail. Shims can considerably reduce the assembly stress of joint structures and improve the joint stiffness and load capacity, and this effect is more remarkable with the increase of gap values. Liquid shims improve the joint stiffness due to its cohesive ability, while solid shims improve the joint load capacity. Hybrid shims possess a composite shimming effect of liquid and solid shims. Whatever the shim material is applied, the joint stiffness and strength drop with the growth of shim thickness, so strict deviation control method should be taken to ensure the assembly gaps as small as possible. The research results enhance the knowledge of shimming effect on joint structures, and thus offer positive guidance for practical application in aircraft assembly.

Due to small production volume in aircraft industry, the available information of variation sources is often not enough to make assumptions on their probabilistic characteristics, especially in the stage of prototype manufacturing. To deal with the problem, an assembly variation modeling and analysis method based on the elasticity mechanics and interval approach is proposed for aircraft assembly. First, variation sources are modeled as bounded convex sets, which are defined as interval structural parameters in interval arithmetic. Then, variation modeling and analysis are successively implemented using the method of influence coefficient and interval arithmetic. After that, a uniform-splitting method is applied to achieve the refinement of the interval extension in variation analysis. To reduce the complexity of the finite element analysis and assembly variation computation, part deformation forms including warpage and torsion are concisely characterized with angle instead of the deviations of isolated key points on the part. The comparison of the assembly variations estimated with the proposed variation analysis method and actual experiment results verifies the effectiveness of the constructed assembly variation model and the proposed method. The interval approach–based assembly variation analysis method is a good complement to traditional probabilistic approach–based methods for compliant assembly systems, which is suited for linear and linearized nonlinear assembly systems. The proposed method provides an improved understanding of the application of compliant assembly variation analysis methods in aircraft manufacturing.

The accuracy of industrial robots is critical in many manufacturing applications. In this case, a self-developed automated fiber placement heavy-duty robot is used for the lay-up of carbon fiber composite prepregs, which requires high positioning accuracy. The mass of the robot body and end-effector exceeds 3 tons and 1 ton, respectively, resulting in considerable deformation and other errors due to the huge size and mass. Paradoxically, the work that our robot engaged in requires high precision, and the positioning accuracy needs to be less than 0.5 mm. This paper conducts a comprehensive analysis of the robot joint error, so as to improve the accuracy of the robot. A torsional and capsize deformation model for joints is established. In addition, the capsize torque and capsize axis of joint deformation are derived to determine the transformation relationship between joint frames. Chebyshev polynomials are used to describe joint-dependent errors. The Levenberg–Marquarelt (L–M) algorithm was used to identify the error model parameters. The effects of different error factors on the positioning accuracy were compared and analyzed. The validation result shows that the proposed model describes 88.11% of the positioning error, and the average residual error of the calibration can reach 0.132 mm. The identified joint-dependent error and the compliance error are 0.296 mm and 0.240 mm, accounting for 26.67% and 21.62% of the positioning error, respectively.

Dimensional variation in aircraft panel assembly is one of the most critical issues that affect the aerodynamic performance of aircraft, due to elastic deformation of parts during the positioning and clamping process. This article proposes an assembly deformation prediction model and a variation propagation model to predict the assembly variation of aircraft panels, and it derives consecutive three-dimensional deformation expressions which explicitly describe the nonlinear behavior of physical interaction occurring in compliant components assembly. An assembly deformation prediction model is derived from equations of statics of elastic beam to calculate the elastic deformation of panel component resulted from positioning error and clamping force. A variation propagation model is used to describe the relationship between local variations and overall assembly variations. Assembly variations of aircraft panels due to positioning error are obtained by solving differential equations of statics and operating spatial transformations of the coordinate. The calculated results show a good prediction of variation in the experiment. The proposed method provides a better understanding of the panel assembly process and creates an analytical foundation for further work on variation control and tolerance optimization.

Robotic boring is an effective way to implement finish machining of intersection holes. However, to a certain extent, its application is limited due to the low rigidity of the robot, whose stiffness brings on high vibration levels. In this study, a new approach based on an equivalent stiffness is proposed to gain a fundamental understanding for the cutting mechanism and vibration performance of a robotic boring system. In the approach, the robotic boring system in one direction is regarded as a mass–spring–damping unit according to the structure characteristics of the robot. Thus, the whole robotic boring system is equivalent to a mass–spring–damping group in three-dimensional space. The stiffness and natural frequency of the robot system were measured by stiffness identification and a modal test on an ABB IRB 6600-175/2.55 robot. An equivalent three-dimensional finite element model based on this approach was established to simulate the robotic boring process, and a verification experiment was conducted to determine the accuracy of the finite element simulation. The results show that the simulated cutting force components and the amplitude in the feed direction are in good agreement with the experiment under different cutting conditions, and this proposed approach is feasible.

A large aircraft fuselage panel is commonly composed of a variety of thin-walled components. Most of these components are large, thin and compliant, and they are also prone to some flexible deformation during assembly and remain deformed after assembly. Besides, many different fabrication and assembly manners are adopted in order to guarantee the complicated assembly relationships between each component. The above characteristics often cause large aircraft fuselage panels to exhibit low stiffness and weak strength, thereby inducing deformation during assembly. Since the posture of a large aircraft fuselage panel is commonly evaluated by matching the theoretical and actual positions of the measurement points placed on it, and its assembly deformation is also represented by the position errors of the measurement points, a reasonable measurement point placement is significant for the large aircraft fuselage panel in digital assembly. This article presents a method based on the D-optimality method and the adaptive simulated annealing genetic algorithm to optimize the placement of the measurement points which can cover more deformation information of the panel for effective assembly error diagnosis. By taking the principle of the D-optimality method, an optimal set of measurement points is selected from a larger candidate set through adaptive simulated annealing genetic algorithm. As illustrated by an example, the final measurement point configuration is more effective to maximize the determinant of the corresponding Fisher Information Matrix and minimize the estimation error of the assembly deformation than those obtained by other methods.

For automated fiber placement (AFP), it is crucial to research the contact characteristics between tow and roller during the tow feeding process. However, it has not been widely explored. In this paper, a theoretical model of contact characteristics for the tow feeding process is established by analyzing the distribution of normal stress, tangential stress, adhesion zone, and slip zone, respectively. The model proposed a numerical algorithm for calculating the contact characteristic. Then the contact characteristic of the tow feeding system can be acquired and validated by the experiment and finite element method. The model that provides theoretical support for AFP optimization. Furthermore, proper pressure (P = 0.6Mpa) and tension (T = 3 N) for towpreg feeding under experimental conditions were obtained.

Large components such as wing sections should be aligned and positioned in a desired position prior to the final manufacturing or assembly, so a digital alignment and position device based on six degree of freedom parallel manipulator with several prismatic-prismatic-prismatic-spherical branches was designed and fabricated to meet the needs. The digital alignment and position device is an alignment device based on the parallel manipulator, and it is also a positioning and holding fixture for large components manufacturing or assembly. In order to effectively and efficiently select the supporting points of large component where the digital alignment and position device branch is connected with, the performance of the parallel manipulator and the fixture should be comprehensively analyzed at the same time. The global performance indices such as the dexterity index, the bearing capacity index, and the stiffness index are calculated based on the mechanism Jacobean matrix of the parallel manipulator, and the positioning stability index is calculated based on the position Jacobean matrix of the fixture. The results show that the global performance indices are not related to the pose, but the positioning stability index is; in addition, all indices rely on the originally selected supporting points and provide the basis for the effective selection of the supporting points for large component aligned and positioned by the digital alignment and position device, which is based on parallel manipulator with prismatic- prismatic-prismatic-spherical pairs.

A large fuselage panel is always connected to a full-scale shape-preserving bracket to enhance its stiffness in digital assembly processes and reduce assembly deformation induced by factors such as gravity. However, the full-scale shape-preserving bracket’s characteristics, such as large size and weight, usually result in a complex installation process and high maintenance costs. A method for optimizing a multi load-transmitting device based discrete support layout is presented in this paper. By constructing finite element model including the fuselage panel and LTDs, the panel’s strain energy under different support layout parameters is obtained according to the principle of mixed uniform design. Using the method of partial least squares regression, the mathematical relationship between the strain energy and the support layout parameters is established, which is then used to obtain the optimum multi LTD based support layout. Furthermore, by comparing the panel deformation under the optimum multi LTD based support with that under the full-scale shape-preserving bracket-based support, we conclude that the optimum multi LTD based support can not only suppress panel deformation effectively, but also keep the panel’s stiffness comparable to that under the full-scale shape-preserving bracket-based support.

The interference fit can improve the fatigue performance of mechanical joints and is widely used in aircraft assembly. In this paper, specimens of lap plates and several interference fit sizes were designed, and then the interference fit hi-lock bolt insertion was carried out in an experimental test. Using the commercial finite element software ABAQUS, a two-dimensional axisymmetric finite element model was established to simulate the bolt insertion process. The finite element model was validated by comparison of experimental results and finite element prediction for insertion force and protuberance height. After the interference fitted bolt insertion, the changing characteristics of the non-uniform hole expansion and protuberance were presented with increases in interference fit size. Under low level of interference fit, the tensile hoop stress was produced mainly on the hole wall, and changed into compressive hoop stress when interference fit size is larger. The maximum tensile hoop stress point on faying surfaces went away from the hole wall with interference fit size increasing.

Intersections and discontinuities commonly arise in surface modeling and cause problems in downstream operations. Local geometry repair, such as cover holes or replace bad surfaces by adding new surface patches for dealing with inconsistencies among the confluent region, where multiple surfaces meet, is a common technique used in CAD model repair and reverse engineering. However, local geometry repair destroys the topology of original CAD model and increases the number of surface patches needed for freeform surface shape modeling. Consequently, a topology recovery technique dealing with complex freeform surface model after local geometry repair is proposed. Firstly, construct the curve network which determine the geometry and topology properties of recovery freeform surface model; secondly, apply freeform surface fitting method to create B-spline surface patches to recover the topology of trimmed ones. Corresponding to the two levels of enforcing boundary conditions on a B-spline surface, two solution schemes are presented respectively. In the first solution scheme, non-constrained B-spline surface fitting method is utilized to piecewise recover trimmed confluent surface patches and then employs global beautification technique to smoothly stitch the recovery surface patches. In the other solution scheme, constrained B-spline surface fitting technique based on discretization of boundary conditions is directly applied to recover topology of surface model after local geometry repair while achieving G1continuity simultaneously. The presented two different schemes are applied to the consistent surface model, which consists of five trimmed confluent surface patches and a local consistent surface patch, and a machine cover model, respectively. The application results show that our topology recovery technique meets shape-preserving and G1continuity requirements in reverse engineering. This research converts the problem of topology recovery for consistent surface model to the problem of constructing G1patches from a given curve network, and provides a new idea to model repairing study.

In the manufacturing engineering field, a multi-robot coordination platform is often used to adjust a large subassembly's location (including position and orientation). This paper introduces an open spherical joint coupling between the subassembly and every robot manipulator and presents a B-spline interpolation trajectory planning method for the manipulators. The location of the subassembly can be described with an element of the group of spatial displacement in Euclidean space. The trajectory of every manipulator can be obtained through finding a smooth motion of the subassembly and then mapping it to the corresponding axes of every manipulator. Unit quaternion is adopted for describing a given set of orientations, and then a cubic B-spline interpolation method is applied to calculate the position and orientation curves, respectively. Finally, the two curves are fused. This paper proves that all the motion points of the subassembly are continuous in the same order as the B-spline function.

This paper presents a novel modeling and compensation method for the volumetric errors of a six-axis gantry automated fiber placement (AFP) machine. Based on the screw theory, the forward and inverse kinematics models of the AFP machine are established. In order to improve the accuracy of the inverse kinematics solution, the Paden-Kahan sub-problem method is used to perform the inverse kinematics solution for the simplified topology of the rotary axes. Using error motion twist to establish a volumetric error transfer model for 54 geometric errors. According to the measurement data of a laser tracker and the Levenberg-Marquardt method to identify the geometric error parameters. The explicit formula of the inverse kinematics solution is used to obtain the error compensation of each motion axis, and the G code of the laying path is modified by the iterative method to realize the compensation of the volumetric errors. By comparing the positions of the tool center point before and after the error compensation, the practicability of the volumetric error modeling and iterative compensation methods are verified, and the geometric accuracy of the AFP machine can be effectively improved.

Through-thickness reinforcement is a promising solution to the problem of delamination susceptibility in laminated composites. Modeling Z-pin–prepreg interaction is essential for accurate robotics-assisted Z-pin insertion. In this paper, a novel Z-pin insertion force model combining the classical cohesive finite element (FE) method with a dynamic analytical fracture mechanics model is proposed. The velocity-dependent cohesive elements, in which the fracture toughness is provided by the analytical model, are implemented in Z-pin insertion FE model to predict the crack initiation and propagation. Then Z-pin insertion experiments are performed on prepreg sample with metallic Z-pins at different velocities to identify the analytical model parameters and validate the simulation predictions offered by the model. Dynamics of Z-pin interaction with inhomogeneous prepreg is described and the effects of insertion velocity on prepreg contact force are studied. Results show that the force model agrees well with experiments and the fracture toughness rises with the increasing Z-pin insertion velocity.