International Journal of Precision Engineering and Manufacturing最新文献

In this study, we conducted an analysis and evaluation of the heat affected zone (HAZ), which serves as a measure of surface processing quality in laser machining of Carbon Fiber Reinforced Plastics (CFRP). Carbon fibers have two axes, horizontal or vertical, in the alignment direction of the fibers. When the optical energy of the laser is primarily conducted along the alignment direction of the carbon fibers and diffused into heat, the HAZ occurs on the surface of the CFRP, exhibiting anisotropic characteristics. In laser processing by pulse, the accumulation of residual heat energy within the carbon fibers induces evaporation and thermal deformation of the polymer at the carbon fiber boundaries, ultimately resulting in a permanent change in the properties of the CFRP, defined as the HAZ. To confirm the influence of process variables on HAZ formation in laser machining, ray tracing was applied to predict the thickness and length of the layer in which the laser beam is absorbed by the CFRP. Generally, it was confirmed that more than 90% of the laser beam is absorbed by three layers of fibers from the surface. Based on this, the temperature distribution of carbon fibers due to residual laser energy during laser machining was predicted. Through these results, the size of HAZ according to the arrangement direction of carbon fibers could be numerically predicted. Experimental results confirmed that process variables such as laser power density and scan speed affect the formation of HAZ. Additionally, the size of HAZ due to conduction along the arrangement direction of carbon fibers was experimentally verified, and quantitative comparison and analysis were conducted with numerical results from previous modeling. Through this analysis, it was possible to predict the size of HAZ affecting surface quality during laser machining of CFRP and validate optimized laser process variables.

Flexures are commonly used in various fields, including optics, microscopy, robotics, and precision engineering since they offer several advantages over traditional motion systems that employ mechanical bearings or sliding mechanisms such as high precision, minimal friction and hysteresis, lack of backlash, compact and lightweight, no lubrication and high stiffness. Flexure may experience cross-coupling effects, which occurs when motion along one axis affects motion along another axis due to mechanical interactions between the flexures. This can lead to unwanted motion errors of flexures or flexure-induced motion errors. This paper discusses identification and compensation of the flexure-induced motion error with a voice coil motor. Firstly, we built one DOF motion system supported by flexure and drove the motion system with a voice coil motor. Flexure-induced motion errors are then identified experimentally using single sine excitation. The single sine excitation can quantitively estimate the inclination angle of the flexure motion system, which makes cross-coupling error. Finally, a feedforward compensation method for the flexure-induced motion error with the voice coil motor is proposed and verified through experimentation.

The quadrotor suspended payload systems have difficulties to achieve precise positioning of the quadrotor UAV and suppression of the load swing angle at the same time because of their characteristics, such as underactuated, strongly coupling, nonlinear and so on. An enhanced coupling anti-swing controller is proposed to solve these problems. With analysis on the relationship between the displacement of quadrotor UAV and the payload swing angle, the three-dimensional dynamic variable rope length model of a quadrotor UAV based on Euler–Lagrange equation is built. An anti-swing method based on energy coupling is designed for this system to realize positioning and eliminate the problem of residual load swing. A new auxiliary signal function is constructed to enhance the coupling relationship between UAV position and load position. The control strategy of three-dimensional quadrotor UAV is designed based on the signal function by the Lyapunov stability theory. The stability of this system is proved by the Lyapunov method and the Barbalat's Lemma. With the experiment, the proposed method is compared with different control methods, the results show that the proposed method can realize the accurate positioning of the quadrotor UAV more quickly and restrain the load swing more effectively. It has strong robustness and stability after the system is added to the disturbance.

Carbon fiber honeycomb composites have been widely used in various aerospace industry due to its superior mechanical properties. However, machining honeycomb composites remains a challenging task due to its difficult-to-machine properties, which can lead to serious machining issues like debonding of the honeycomb wall and burrs. Cutting angles is a key factor affecting the damage formation of the carbon fiber honeycomb composites. In order to reduce machining damages, in this study, an analytical method for calculating the cutting angles was proposed, in which the structural characteristics of the honeycomb cell and the interaction between the honeycomb wall and the cutting tool were considered. And deformations of the honeycomb wall with different configurations of cutting angles were analyzed. Then cutting experiments were conducted with different cutting directions and cutting width, which affect the cutting angles. Finally, the effects of cutting angle on the cutting forces and machining quality are discussed. The results show that larger in-plane cutting forces are beneficial to decrease the burrs of the honeycomb wall, and the quality of the machined honeycomb wall is poor when the cutting angles range from about 66° to 141°.

This study aims to develop an artificial intelligence-based model for analyzing the condition and detecting anomalies by encoding time-series data from manufacturing processes as images. Deep learning has demonstrated the significance of data analysis and anomaly detection in the vision field, and Convolutional Neural Networks (CNN) models have shown exceptional performance and high applicability in image analysis. Based on this, our study intends to utilize image encoding techniques to perform anomaly detection on time-series data. Data such as force, vibration, and sound from equipment during the manufacturing process are collected and transformed into images using various methods, including Gramian Difference Angular Field, Gramian Summation Angular Field, Markov Transition Field, and Recurrence Plot (RP). The transformed image data is then trained and classified for equipment conditions using various CNN models. Finally, we adopt the RP image encoding method and ResNet50 model, which demonstrated the highest accuracy of 99.6%, and compare them to the top 5 models. Based on the high accuracy demonstrated by the top five models, our proposed approach has proven to have significant performance, exhibiting a high success rate of over 90% even when applied to actual data for CNC-machining process. Through this, we propose a process that utilizes the explainable AI Grad-CAM system to identify the feature layer area of the image and confirm the presence of anomalies. With the proposed process, workers can identify abnormal areas or segments of abnormal conditions in the transformed image graph. By providing evidence for state judgment, even inexperienced workers can easily check the condition of manufacturing equipment.

Linear guides are essential components of precision equipment, such as computer numerical control machine tools and coordinate measuring machines. The key to improving their accuracy is to conduct high-precision measurements and effectively compensate for the six degrees of freedom errors that emerge during motion. However, the roll is a difficult parameter for achieving high-precision measurements in laser measurements, and it has yet to be effectively addressed. This study presents a high-precision and high-sensitivity roll measurement method based on the laser polarization state. The method employs a quarter-wave plate as the roll angle sensitive component. Firstly, the feasibility of this method was assessed through the Jones matrix and Zemax simulations. Subsequently, both theoretical and experimental analyses were conducted to evaluate the impact of using a corner cube reflector as a retroreflector on roll measurements. Finally, a corresponding measurement device was developed, and a series of comprehensive experiments were executed. The experimental findings indicate a measurement sensitivity of 7 mV/arcsec for this method. In comparison to commercial autocollimators, the contrast deviation within ± 100 arcsec is ± 1 arcsec. The repeatability of measurements on the air-floating guide rail is 1.3 arcsec, with a maximum contrast deviation of 2.5 arcsec compared to the electronic level. Thus, the proposed method offers the advantages of high precision, simple structure, and low cost and provides a new technique for high-precision roll measurement.

Renowned for its non-contact nature, high-resolution capabilities, and suitability for real-time control, ECDS (eddy current displacement sensor) operates on electromagnetic induction principles to detect even the minutest changes in the position of conductive targets. The interface between an ECDS and external instrumentation stands as a critical juncture influencing signal conditioning, noise reduction, compatibility with modern data acquisition systems, and adaptability to dynamic measurement environments. This paper presents a novel direct digital interface of frequency-modulated (FM) ECDS for real-time control of AMB. First, an FM eddy-current gap sensor is built, and its LC oscillation frequency variation is detected with the proposed direct digital interface. The direct digital interface has several distinct advantages such as simple configuration, noise immunity, and adaptability. In particular, both the DSP's internal SW (software) prescale and an external HW (hardware) prescale, such as a counter are used together to balance the static and dynamic performances, which offers moderate resolution and sufficient stability for real-time applications. Finally, both the static and dynamic performances of the FM eddy-current gap sensor are investigated by measuring the levitation jitter and closed-loop sensitivity function of the one-DOF AMB system.

This paper reviews earlier studies focusing on thickness measurements of thin films less than one micrometer thick. Thin films are a widely used structure in high-tech industries such as the semiconductor, display, and secondary battery industries. Typical non-destructive and non-contact techniques for measuring the thickness of thin films are spectral reflectometry (SR) and spectroscopic ellipsometry (SE). SR can measure the thin-film thickness with a simple layout. With the combination of SR and optical interferometry, the simultaneous measurements of thin film and 3D surface profiles or thick layer have been proposed and demonstrated. For an analysis and verification of SR, several works including artificial intelligence algorithms and uncertainty evaluations have been published. SE can measure thinner thicknesses with more information pertaining to the polarization state, incident angle, wavelength, and etc. According to the type, location, and number of elements that make up the basic optical layout, ellipsometers can be classified into five types. Based on a mathematical model of the ellipsometric transfer quantity, the operational principle and measurement procedure are discussed. To ensure measurement reliability, the uncertainty components of the SE were evaluated. With the development of high-tech industries in the future, thin-film thickness measurement techniques can be expected to find wider use with faster measurement speeds, a higher dynamic range, and better measurement reliability.

The rising applications of hybrid composites demand the best manufacturing quality solutions. This article uses a newly developed hybrid composite material of carbon, basalt and Kevlar-29 woven fibres manufactured for laser drilling operation for its characterization and optimization using a golden jackal optimizer. The Nd: YAG laser parameters used are lamp current (LC), pulse width (PW), stand-off distance (SoD) and compressed air pressure (CAP). The responses considered for this work are hole taper angle (HT) and average overcut (AvgOC). The experimental set was designed through the Box–Behnken design. Optimization has been carried out using the Golden Jackal optimizer. Optimized responses are theoretically obtained and validated with the confirmation experiment as LC = 220 A, PW = 3 ms, SoD = 1.1 mm and CAP = 12 kg/cm². Whereas predicted and experimental values of HT and AvgOC are recorded as 0.113° and 0.125° and 0.066 and 0.098 mm, respectively. The confirmatory experiment yielded a good agreement with the predicted responses.

The primary challenge in CNC machining is optimizing arc interpolation for better smoothness and precision, as traditional methods often fail to maintain a consistent curvature, resulting in inefficiencies and inaccuracies. This study utilizes the Archimedes spiral, comprising a series of seamlessly connected circular arcs, as an innovative interpolation curve to improve arc interpolation efficiency and accuracy. The proposed methodology addressed the issue of constant arc interpolation curvature and facilitated the implementation of the spiral’s interpolation algorithm. It employs the parametric equation of the Archimedes spiral to define tangent vectors, which is pivotal for the execution of the algorithm. In the context of speed planning, the curve was segmented into different sections, allowing the calculation of maximum acceleration based on the interval segmentation angle of the tangent vector. This segmentation facilitates the analysis of speed variations, which are consequently integrated to determine the speed distribution across different curve sections. Through this integration, the motion process is categorized, thereby achieving a refined speed distribution curve. This study introduces a realtime interpolation algorithm capable of calculating the largest axis of the tangent vector, thereby enabling precise pulse coordinates. This coordinate information are then effortlessly derived from the simple geometric relationships inherent in the spiral's structure. The effectiveness of the proposed method is demonstrated through simulation and practical machining on two distinct graphical representations. Compared with the traditional linear interpolation algorithm, the proposed method improves the machining efficiency of heart-shaped curves by 6.85% while ensuring the machining accuracy. Comprehensive evaluation, encompassing track error, and surface roughness assessments, validates the enhanced performance of this interpolation technique.