Sheet metal bending, an essential part of manufacturing, is increasingly used in transportation, aerospace, and related fields. Challenges related to forming accuracy and path generation have always been major concerns in industrial applications. A prediction model based on GA-BPNN (Genetic Algorithm-Backpropagation Neural Network) was developed, determined, and optimized to mitigate the impact of springback on the forming angle. A 9-input model, considering material classification, was adopted to further enhance prediction accuracy. An artificial potential field (APF) method based on discrete geometric calculations of main and auxiliary (DMA) points was introduced to improve the reliability of a collision-free bending path for the shape characteristics of sheet metal parts, and further optimized using the rapidly exploring random tree (RRT) method. The results indicated that the maximum errors in forming angles were reduced from 2.45 to 0.4°, and 0.15°, respectively, when applying different models. The experimental collision-free path maintained a safe distance of 2 mm, ensuring that the robotic arm operated smoothly to the target position. Additionally, this study could advance the development of a more accurate and efficient bending process in industrial settings.

Aiming at the issues of poor dimensional consistency,low profile,and wall thickness forming accuracy of the complex cabin intake port surface,a tolerance-constrained adaptive machining approach for negative allowance surface based on the composite on-machine measurement is proposed.A composite on-machine measurement platform is constructed with the hysteresis effect clarified,and a synchronous data acquisition system is developed.The measurement error is then examined along with a proposed iterative calibration method for laser sensors.Under the guidance of measured profile data,the high-precision ultrasonic on-machine measurement of wall thickness is achieved by accurately solving normal vectors.For machining surfaces with negative allowance,a mathematical model is developed under double constraints of profile and thickness.Additionally,occupancy,non-uniformity,and consistency are defined,and under various weights,a data-driven surface reconstruction is.realized.Experiments are used to develop and validate an on-machine measurement and adaptive machining strategy for the intricate cabin inlet surface.The results demonstrate that the positive and negative errors of the measured profile are+0.0083 and-0.0073 mm,respectively,and the errors of wall thickness are within 0.03 mm.When the surface has a negative allowance of-0.0524 mm,the profile's maximum positive and minimum negative deviations after adaptive machining are+0.0538 and-0.3655 mm,respectively.And the maximum and minimum values of the wall thickness deviation are+0.16 and-0.22 mm,respectively,which satisfy the tolerance requirements and achieve accurate and effective machining of the whole surface.

Laser Triangulation On-Machine Measurement (LTOMM) is being implemented increasingly to inspect aeronautical components accurately and efficiently, with its enhanced application in adaptive machining. This work proposes an error compensation and controlling method for measuring the typical features of steps, holes, and freeform surfaces to improve accuracy. Then, the global path to inspect the cabin’s structures is planned by introducing optimization algorithms, thus providing an appropriate sequence to shorten the traveling length. After these, the test piece was designed, measured, and manufactured using the adaptive machining process that integrates the LTOMM. The results show that the measurement errors of steps, holes, and freeform surfaces are +0.0092, −0.006, and +0.0406 mm, respectively, and further reduced to +0.0013, −0.0019, and +0.0083 mm after error controlling. The cabin’s freeform surface was fabricated with the maximum positive and minimum negative errors of +0.184 and −0.123 mm, which is evaluated by the mechanical probe. The measured data-driven machining process can guarantee that the error satisfies the required tolerance, promoting the application of the LTOMM process in aeronautical intelligent manufacturing.

Laser triangulation on-machine measurement (LTOMM) with associated processing technology is an effective way to evaluate freeform surfaces' quality without removing the workpiece from the machine tool. In this article, an integrated system able to acquire the sensor's one-dimensional displacements and three-dimensional locations is presented. This system is able to prevent spatial mismatch, and then a demand-oriented LTOMM process is proposed. Different from conventional understanding of the trade-off behavior between accuracy and efficiency, an error management strategy is formed to fulfil the required accuracy by precisely controlling the inclination angle and measurement displacement. To improve efficiency, these parameters are fine-tuned adaptively to the surface characteristics by an accuracy-monitoring path optimization algorithm (AMPOA) developed. The unevenly distributed peak error is suppressed by utilizing a mathematical model to regulate the measurement path without compromising efficiency. A simulation is conducted to substantiate that this method is not only able to improve measurement efficiency but also to suppress the peak error. The algorithm and the model were verified on a freeform turbine blade surface with a required accuracy of 20 mu m. The efficiency was improved by 42.45%, and the peak error was reduced. This would help both the researchers and industries with a more efficient and reverse-controllable demand-oriented LTOMM approach under the constraint of required accuracy in the future.

The issue of springback in sheet metal bending has always been of great concern in industrial applications. To reduce its influence on forming angle in bending process., a prediction model based on GA-BPNN (Genetic Algorithm-Backpropagation Neural Network) was proposed and optimized. To train the network., the finite element method was introduced to construct the dataset. Then the parameters in the neural network were determined., and a 5-dimensional input improved model considering material classification was adopted to further improve the prediction accuracy. The experimental results showed that the maximum errors of forming angle reduced from $2.45^{\circ}$ to $0.4^{\circ}$ , and $0.15^{\circ}$ by applying different models.

Complex surface plays a vital role in the aeronautical industries for its preferable mechanical properties. However, machining such complexity is challenging without post-process inspection. Compensation derived from measurement is essential to ensure final form accuracy. The process efficiency can be improved significantly with real-time on-machine measurement. The laser triangulation technology is an excellent fit for such applications, because of its micrometric accuracy with a large measuring range and non-contact nature. In this paper, the recent research progress of laser triangulation on-machine measurement (LTOMM) has been reviewed. The measuring technique, error compensation, path optimization, and application of LTOMM are reviewed. More specifically, the possible errors in laser triangulation are reviewed and comprehensively analyzed based on the recent research published. The measurement path planning and optimization strategies are surveyed. An application is presented to illustrate the effectiveness and efficiency of LTOMM in propeller or turbine blade fabrication. Finally, the challenges, development trends and perspectives of laser triangulation on-machine measurement are discussed.

The core of graphical programming for sheet-metal bending is operation planning based on parameter identification, which can directly determine the bending quality and efficiency. Given this, the proposed graphical programming system and method can predict the inner arc radius of bending oriented to programming demands and optimize multi-step bending operations. At first, a neural network model based on multi-layer perceptron (MLP) was established, considering the number of neurons and hidden layers. In the bending process a fitness function was defined, and the discrete particle swarm optimization (DPSO) was improved using the adaptive inertia weight and the genetic algorithm (GA). Finally, the operation planning of 10-step and 20-step bending was simulated, and bending tests were conducted on sheet-metal workpieces for verification. The results show that the algorithm’s maximum positive error and minimum negative error in predicting the arc radius are 0.12 mm and − 0.16 mm, respectively. When there are 10 bending steps, the particle swarm optimization-genetic algorithm (PSO-GA) converges after only ten evolutions; when the bending steps increase to 20, the fitness value finally stabilizes at 1.31. The optimal operations of the experimental six-step bending part are 6 → 5 → 4 → 3 → 1 → 2. This result indicates that the workpiece is turned over three times and turned around once in the sheet-metal bending process, which is consistent with the simulation, thus verifying the bending efficiency and accuracy.

In the process of sheet metal bending, the assistant robot is widely used for feeding. However, due to the low absolute positioning accuracy of industrial robots, the pose error of sheet metal in the process of feeding is inevitable, which has a great impact on the accuracy of bending. In order to improve the positioning accuracy of sheet metal, a dynamic Elman neural network-model optimized by sparrow search algorithm(SSA-Elman) is proposed in this paper for accurate prediction and compensation of the pose error. Sparrow search algorithm has strong optimization ability and fast convergence speed. It can overcome the shortcomings of slow convergence speed and easily falling into local optimal value of standard Elman neural network. Therefore, sparrow search algorithm can be introduced to correct the prediction error of Elman neural network. Finally, the data prediction effect of SSA-Elman model is compared with that of the non-optimized Elman neural network model, and four evaluation indexes of the mean absolute error(MAE), the mean square error(MSE), the root of mean square error(RMSE) and determination coefficient are used for estimation. Results show that the mean absolute error, the mean square error and the root of mean square error of SSA-Elman model are reduced by 71.5%, 92% and 72.9% respectively, and the determination coefficient is increased by 10.2%, which indicate that SSA-Elman model has higher prediction accuracy.

In order to solve the problems of low efficiency and high collision probabilities of traditional bending, the collision between the sheet metal parts and the bending machine cannot be ignored upon establishing a coordinate system. Considering the factors such as pose calculation when moving parts, an improved Artificial Potential Field method based on discrete geometric calculation of main and auxiliary points is proposed for the shape characteristics of sheet metal parts. The potential field path controls the motion path of the sheet metal part and finds a collision-free bending robot working path. Finally, the improved method is simulated, and the results show that the algorithm can successfully guarantee no collision between the sheet metal part and the bending mechanism, and ensure that the robotic arm runs smoothly to the target position.

In-haul cables must be examined and maintained regularly to ensure the smooth operation of cable-stayed bridges. The magnetic flux leakage (MFL) detection method is being applied increasingly commonly to inspect defects. For the quantitative identification of the broken wire, we designed an MFL inspection equipment and investigated the performance. A trapezoidal magnetic dipole model is constructed to simulate the leakage distribution, and the axial and circumferential excitation schemes are used to obtain the intensity of magnetic induction. Then, based on the simulation of model parameters, an MFL detection system is configured for detecting the damage in the broken wire. To investigate the adaptability of the developed system, a rectangular defect is inspected in the vertical and spiral climbing modes. The signal acquired under different paths is processed by the median filtering and wavelet transform methods and analyzed. A random forest algorithm is used to quantitatively identify the number of broken wires, defect dimensions, and cross-sectional area loss to verify the proposed method. The results show that the axial excitation generates a single peak signal, which offers better detection of the wire defects. In the vertical climbing mode, the maximum detection errors in the width and cross-sectional area loss are 0.64 mm and 0.46%, respectively, while the values are 0.21 mm and 0.1% in the spiral climbing mode, indicating that the latter mode offers higher identification accuracy. Furthermore, in future work, the detection capability of irregular defects should be studied, expanding its application in health monitoring.

On-machine measurement is being used increasingly to inspect the features of machined components without removing the workpiece from the machine tool. In the present work, a laser displacement sensor is integrated on a machine tool to measure the elementary features of a step, hole, cylinder, and plane. An iterative calibration method is introduced to obtain the position and direction of the sensor accurately, and data processing is performed for filtration and edge recognition. The accuracy and uncertainty of detecting the characteristic parameters of height, diameter, roundness, and flatness by the laser-based on-machine measurement are evaluated and are compared with those of a mechanical probe. Furthermore, the capability of the on-machine measurement process on the adopted machine tool is assessed according to ISO 22514-7. The results show that the difference between the calibrated and actual sphere center is reduced to 4.7 μm after five iterations. The on-machine measurement by the mechanical probe is more accurate, while that by the laser displacement sensor is more efficient. The uncertainties of the two methods are evaluated as being in the ranges of 2.4–3.7 μm and 4.8–5.8 μm, respectively. The accessible tolerances can be obtained to determine whether the machine tool is competent for requirement as a measuring device.

On-machine measurement (OMM) is increasingly used to inspect freeform surfaces without unloading the workpiece from the machine tool. Compared to the mechanical probe deployed in many OMM processes, a laser displacement sensor has its advantage of high scanning efficiency. However, the accuracy has been a bottleneck limiting its application in precision inspection, due to the lack of a rigorously planned path. In this work, an error compensation model is established and verified considering laser triangulation error, misalignment from calibration, tooling error, and preevaluated error. Essentially, error modeling contributes to improve the measurement accuracy of laser triangulation OMM (LTOMM). After understanding these errors, a reliable measuring path is crucial to maintain the given accuracy of LTOMM. This path generated is capable of minimizing the measurement time at a given accuracy constraint. In a particular case study, a turbine blade surface was inspected using LTOMM following generated path against a built-in probe. The measurement time was reduced by 77.4% for a constraining accuracy of ±10 μ\textm. The time could be further reduced by 38.6%, for a constraining accuracy of ±13 μ\textm. This study has demonstrated that LTOMM could potentially replace the on-machine probe during measurement of freeform surfaces, toward more efficient industrial applications in the future.

轴类零件主要用来传递转矩和承受载荷,常通过数控车床加工而成其精度要求高,可用激光位移传感器对其轮廓信息进行采集,但由于机床存在位置及运动误差等,使得在机测量精度较低.因此为提高测量精度,提出面向阶梯轴类零件激光测量的实时误差补偿算法,并分别对补偿前后台阶测量点云进行模型重构,对比分析其测量误差.对比分析结果验证了补偿算法的有效性.台阶面测量表明,经几何关系修正后,其平面测量误差与三坐标测量结果差值为0.002 mm,证明测量拟合精度高.

Faced with many turned parts on a lathe, the profile inspection usually relies on coordinate measuring machine, which is an off-line method and may introduce repositioning or reclamping errors. Then, an accurate and efficient on-machine measuring method of integrating a laser displacement sensor on a numerical control lathe has been proposed, which can inspect the profile of the workpiece without taking it off from the machine tool. Based on the measuring process, a deviation model is constructed after analysis of the error sources. To improve the accuracy of the measurement, the deviations are identified by several other related inspection methods, which are employed to compensate the measured values. Subsequently, an additional experimental test is conducted to verify the effectiveness of the proposed methodology. Results have shown that the positive and negative deviations in the static measurement are +4.4 mu m and -2.7 mu m, respectively. Then, to improve the measurement efficiency, the dynamic process is adopted, in which the workpiece rotates at the optimized parameters. Although the measuring deviations enlarge to +6.3 mu m and -3.1 mu m partially, the time consumption is reduced to 10 % to that of the static mode. The confidence interval analysis implies that the probability of 95 % of the authentic values will fall into the calculated intervals.

Faced with many turned parts on a lathe, the profile inspection usually relies on coordinate measuring machine, which is an off-line method and may introduce repositioning or reclamping errors. Then, an accurate and efficient on-machine measuring method of integrating a laser displacement sensor on a numerical control lathe has been proposed, which can inspect the profile of the workpiece without taking it off from the machine tool system. Based on the measuring process, a deviation model is constructed after analysis of the error sources. To improve the accuracy of the measurement, the deviations are identified by several other related inspection methods, which are employed to compensate the measured values. Subsequently, an additional experimental test is conducted to verify the effectiveness of the proposed methodology. Results have shown that the positive and negative deviations in the static measurement are +4.4 μm and -2.7 μm, respectively. Then, to improve the measurement efficiency, the dynamic process is adopted, in which the workpiece rotates at the optimized parameters. Although the measuring deviations enlarge to +6.3 μm and -3.1 μm partially, the time consumption is reduced to 10% to that of the static mode. The confidence interval analysis implies that the probability of 95% of the authentic values will fall into the calculated intervals.

On-machine measurement is being applied increasingly commonly to inspect deviations in machining, without removing the workpiece from the machine. A laser-based non-contact method has been introduced to measure complex profiles highly efficiently, but the accuracy of its measurement has not been assessed and compensated for. To improve the accuracy of the measurements of inclined and curved profiles using a laser displacement sensor, this study evaluates and compensates for deviations in the measured values by using compensation strategies. The factors influencing measurement accuracy are analyzed and the phenomenon of secondary reflection is explained on the basis of laser triangulation. For an inclined profile, the strategy whereby angle compensation is carried out followed by displacement compensation is used. For a curved profile, a corrected formula in which the coefficients are calculated according to the measurements is applied. The two patterns of coarse and fine compensation rely on the choice of the correction coefficient determined by the adaptive requirements of assessing accuracy. The results show that the surface roughness, measuring environment, and positional errors are less important than the inclination angle and displacement in affecting the accuracy. For the inclined profile, the measurement accuracy improved to -7.4 mu m similar to+6.7 mu m after compensation. For the curved profile, the deviations decreased by 71.5% and 91.9% compared with the original after coarse and fine compensations, respectively. Furthermore, the compensation strategies can be applied to the measurement of complex profiles in future work. (C) 2019 Elsevier Ltd. All rights reserved.

A hybrid approach for the thickness measurement of complex structural parts on a CNC machine tool is proposed. On-machine probing is firstly carried out to acquire the actual shape data for a part, after which the normal directions of the measurement points are estimated, and finally on-machine ultrasonic inspection is performed to measure the thickness. Calibrations of the on-machine probe and the ultrasonic thickness sensor are conducted to improve the measurement precision. The motion accuracy of the CNC machine tool is pre-evaluated to ensure reliability of the measurement data, and the measuring error is found to be less than 1 μm. A case study of a hollow turbine blade is performed, and the results show that the deviation in the thickness measurement at any measurement point is less than 0.02 mm.

Hybrid machining processes have great advantages on enhancing the machining efficiency and accuracy of complex structural parts. Adaptive machining can answer the transition challenges of poor shape accuracy and bad dimensional consistency in hybrid machining processes. A novel shape-adaptive machining approach via direct spatial deformation of template tool cutter positions is developed, which can solve the problems efficiently and quickly. The framework for shape-adaptive approach is illustrated, in which on-machine measurement, globally spatial deformation and locally spatial deformation are involved. The details of on-machine measurement using a touching probe are present. A volume-based free-form deformation method was used to deform the template tool cutter positions globally. A 3rd-degree Bezier surface is built on which the tool cutter positions were locally projected. Finally, a case study of adaptive machining turbine blade was carried out, the results of which validate the feasibility of proposed measured-driven shape-adaptive machining approach. (C) 2018 Elsevier Ltd. All rights reserved.

采用Ag-Cu-Ti合金粉末在加热温度920℃保温时间8 min条件下真空钎焊镀钛金刚石,运用X射线衍射仪、扫描电镜、能谱仪综合分析界面微观结构、元素分布特征及新生物相形貌.结果表明,镀钛金刚石镀层并不是单一钛的附着层,而是薄层TiC.钎焊过程中,镀钛金刚石磨粒与Ag-Cu-Ti钎料结合界面出现显微组织分层现象,原镀层与新生层厚度分别约为0.8 μm和5μm.新生化合物在镀层表面呈短针状致密生长且有序连成片状,在镀层破损处有长约4 μm,宽约1.5 μm针状物,而在破损边缘处碳化物呈棒状背离镀层侧向生长.原子扩散动力学分析显示镀层晶体结构会限制游离C原子的无规则运动,使得新生TiC均匀分散形核并长大.

The interfacial microstructure of boron‐doped diamond brazed with Ag‐Cu‐Ti alloy , was studied by means of SEM , EDS , XPS . Theory of thermodynamic was applied to analyze the microstructure diffusion mechanism , the structure and grow th process of reaction product . Additionally , the static and impact strength of grain was tested under different conditions . The results showed that in the process of vacuum brazing , TiC was preferentially generated in front of the diffusion . As time went on , TiB2 was formed through combination reaction . Due to the similar thermal expansion coefficients between TiC and TiB2 , composite structure of TiB2‐TiC would be formed and the thermal stress reduced , which resulted in the metallurgical bonding between the grains and the filler . After brazing , the static and impact strength of boron‐doped diamond increased by 5 .4% and 34 .8% , respectively . In addition , with the good self‐sharpening ability , the brazed boron‐doped diamond tool would have a predictable application in difficult‐to‐cut materials .