On-line Inspection Research Articles

Anomaly detection in Skin Model Shapes using machine learning classifiers

The concept of Skin Model Shapes has been proposed as a method to generate digital twins of manufactured parts and is a new paradigm in the design and manufacturing industry. Skin Model Shapes use discrete surface representation schemes, such as meshes and point clouds, to represent surfaces, which makes them enablers to perform an accurate tolerance analysis and surface inspection. However, online inspection of manufactured parts through use of Skin Model Shapes has not been extensively studied. Moreover, the existing geometric variation inspection techniques do not detect unfamiliar changes within tolerance, which could be the precursors to the onset of the manufacturing of out of tolerance part. To detect the unfamiliar changes, as anomalies, and categorize them as systematic and random variations, some unique surface characteristics can be extracted and studied. Random surface deviations exhibit narrow normal distributions, and systematic deviations, on the other hand, exhibit wide, skewed, and multimodal distributions. Using those surface characteristics as key traits, machine learning classifiers can be used to classify deviations into systematic and random variations. To illustrate the method, multiple samples from a truck component manufacturing line were scanned and the collected 3D point cloud data was used to extract features. A prediction score of 97–100% can be achieved by decision tree, k-nearest neighbor, support vector machines, and ensemble classifiers. The purposed approach is expected to extend the existing online inspection approaches and applications of Skin Model Shapes in quality control.

Progress on the ultrasonic testing and laser thermography techniques for NDT of tokamak plasma-facing components

ABSTRACT During manufacturing and operation, different kinds of defects, e.g., delamination or surface cracks, may be generated in the plasma-facing components (PFCs) of a Tokamak device. To ensure the safety of the PFCs, various kinds of nondestructive testing (NDT) techniques are needed for different defect and failure mode. This paper gives a review of the recently developed ultrasonic testing (UT) and laser thermography methods for inspection of the delamination and surface cracks in PFCs. For monoblock W/Cu PFCs of divertor, the bonding quality at both W-Cu and Cu-CuCrZr interfaces was qualified by using UT with a focus probe during manufacturing. A noncontact, coupling-free and flexible ultrasonic scanning testing system with use of an electromagnetic acoustic transducer and a robotic inspection manipulator was introduced then for the in-vessel inspection of delamination defect in first wall (FW). A laser infrared thermography testing method is highlighted for the on-line inspection of delamination defect in FW through the vacuum vessel window of the Tokamak reactor. Finally, a new laser spot thermography method using laser spot array source was described for the online inspection of the surface cracks in FW.

Online Inspection Method of Defect Dimension of Stainless Steel Precision Blanking Sheet Contour

In this study, a new method of online inspection of contour defects is proposed for stainless steel precision blanking sheet to address problems in the use of the current method to obtain detect information: the recognition system is complex, and the measurement error of defect dimension is large. The images of defect blanking sheet are obtained by the established image collection system. The results of image analysis show that the defect contour of blanking sheet has the correspondence and similarity characteristics. An improved sequential similarity detection algorithm and the measurement algorithm are developed. The measurement experiments show that the online inspection system is stable and reliable, and the maximum measurement error is 0.0362mm. The results of this study will lay a foundation for the online inspection of precision machining.

Surface reconstruction of laser-cladding remanufacturing blade using in adaptive belt grinding

The remanufacturing technology of aero-engine blade is one of the effective ways to save expensive materials such as titanium alloy, make full use of resources, and solve the problem of resource hiding. Laser-cladding technology has been widely used in the remanufacturing of blades due to its ability to fabricate high-performance alloy surfaces on metal substrates without affecting the properties of the matrix. In order to solve the problem that the grinding precision is difficult to guarantee due to the deformation and residual distribution of the blade in laser cladding remanufacturing, the self-adaptive grinding process of abrasive belt is repaired based on online inspection and reconstruction model. Firstly, the distribution of theoretical measuring points is planned based on the curvature of cross-section curve, and then the theoretical measurement points are adaptively extracted according to the improved firefly algorithm, so that the CMM measurement path planning of laser cladding and remanufacturing blades is realized. Taking the layer as the unit, the measurement data are processed, and the iterative nearest point algorithm is used for data registration to realize the laser-cladding remanufacturing blade measurement data processing. On this basis, the blade surface model is obtained based on NURBS interpolation. Finally, a laser cladding and remanufacturing blade of an aero-engine was tested. The experimental results show that the blade surface roughness is 0.4 μm, the blade edge accuracy is less than 0.05 mm, and the blade edge error is less than 10%. The experimental results show that this method can effectively solve the grinding process of laser-cladding remanufacturing blade.

Spatially Resolved Acoustic Spectroscopy Towards Online Inspection of Additive Manufacturing

Spatially Resolved Acoustic Spectroscopy Towards Online Inspection of Additive Manufacturing

Application of image edge detection methods for precise estimation of the standard surface roughness parameters: Polypropylene/ethylene-propylene-diene-monomer blend as a case study

Material surface roughness is an important property in different fields of science and its precise measurement is still a serious concern. Roughness measurement is commonly implemented using stylus profilometer. Although it suffers from some drawbacks such as low speed and destructive nature. Optical methods, such as machine vision coupled with image texture analysis have shown promising capability for noncontact/nondestructive roughness measurement. In the present study, a roughness evaluation method is proposed based on image edge detection algorithms. The method was applied to investigate the surface roughness of polypropylene/ethylene-propylene-diene-monomer (PP/EPDM) blend as an important engineering plastic. Different roughness patterns were created on PP/EPDM sheets employing hot press processing. Images of the roughened samples were captured with 10 different resolutions. In the proposed method, the performance of five different edge detectors including Roberts, Prewitt, Sobel, Laplacian of Gaussian (LoG), and Canny were examined. The results showed that LoG method in the images with 200 dpi resolution effectively evaluates PP/EPDM surface roughness. Linear correlation coefficients (R2) between LoG results and Stylus profilometry results was greater than 0.98. Moreover, some mathematical models were developed for evaluation of the roughness parameters based on LoG edge frequency. The models’ results showed 6.7% deviation from stylus profilometry results in the worst case. The proposed method can be used as a feasible solution for roughness evaluation of polymeric materials in online inspections.

Laser backscattering imaging as a non-destructive quality control technique for solid food matrices: Modelling the fibre enrichment effects on the physico-chemical and sensory properties of biscuits

This work focused on modelling the effect of fibre enrichment on the physico-chemical and sensory properties of biscuits by an image analysis of laser backscattering patterns. The study was done on four formulas where, besides the control, three fibre enrichment levels were included (5%, 10% and 20% w/w). The impact on the physico-chemical and sensory properties of biscuits was characterised according to analyses of texture, thickness, area, mass increment, density, flux of solvents (water and milk), colour, taste, mouth texture, etc. Moreover, the image analysis was carried out to collect information from the interaction of a laser with two different biscuit production chain matrices: after the mixing and forming phase (doughs) and after baking (end biscuits). That information was obtained according to the image descriptors generated from the morphology of the observed laser patterns. Both studied matrices offered different, but complementary, information about isolating the variance produced by fibre enrichment for doughs, while the variance generated by heat treatment was also recorded for biscuits. The quantitative prediction of the physico-chemical and sensory properties of biscuits improved when combining both information blocks. The impact of fibre enrichment on biscuits can be modelled by this imaging technique, which could be the basis to develop new non-destructive systems for online inspections made during cookie processing to quickly and non-destructively report physico-chemical and sensory information.

Optimal Wavelength Selection for Hyperspectral Imaging Evaluation on Vegetable Soybean Moisture Content during Drying

Hyperspectral imaging technology is a promising technique for nondestructive quality evaluation of dried products. In order to realize real-time, online inspection of quality of dried products, it is necessary to determine a few important wavelengths from hyperspectral images for developing a multispectral imaging system. This study presents a binary firework algorithm (BFWA) for selecting the optimal wavelengths from hyperspectral images for moisture evaluation of dried soybean. Hyperspectral images over the spectral region 400–1000 nm, were acquired for 270 dried soybean samples, and mean reflectance was calculated from hyperspectral images for each wavelength. After selecting 12 important wavelengths using BFWA, a moisture prediction model was developed using partial least squares regression (PLSR). The PLSR model with BFWA achieved a prediction accuracy of R p = 0.966 and R M S E P = 5.105 % , which is better than those of successive projections algorithm ( R p = 0.932 and R M S E P = 7.329 % ), and the uninformative viable elimination algorithm ( R p = 0.928 and R M S E P = 7.416 % ). The results obtained by BFWA were more stable, with a smaller standard deviation of R p and R M S E P than those of the genetic algorithm. The BFWA method provides an effective mean for optimal wavelength selection to predict the quality of soybeans during drying.

Retraction Notice: An automatic online inspection system for a coupler yoke for freight trains

Retraction Notice: An automatic online inspection system for a coupler yoke for freight trains

Scheduling and Control of a Wafer Transfer Robot for Foundry Equipment Innovation Competition

This paper presents a design and implementation of a wafer transfer robotic system for a Foundry Equipment Innovation Competition. In this system, a 4-DOF robotic arm was designed and constructed for wafer transfer between a cassette and processing chambers. The purpose of this competition was to promote intelligent automation for wafer fabrication, in which multiple wafers and chambers need to be handled simultaneously. A scheduling algorithm is proposed for efficient task execution of wafer to process recipes assigned during the competition. The task also features to integrate defect detection and endpoint determination of wafer processing. The robot was required to receive the processing data from the mission platform to determine the endpoint of wafer processing in a chamber and complete the recipe automatically. The online inspection algorithm detects wafer defects during the task execution and report to the task planner. The state of each wafer and chamber are taken into consideration for scheduling in order to complete the wafer fabrication in an efficient and flexible manner. The developed robotic system spent 9 minutes to complete the assigned mission of transferring 6 wafers among multiple chambers and 2 cassettes and achieved 93.4% defect-detection rate in the competition.

In‐Process Quality Inspection of Rolling Element Bearings Based on the Measurement of Microelastic Deformation of Outer Ring

Quality inspection is the necessary procedure before bearings leaving manufacturing factories. A testing machine with low shaft speed and light radial load condition is generally used to test the dynamic quality of bearings, which avoids creating any potential damages to testing bearings. However, the signal of defective bearings is easily polluted by very weak noise using the traditional vibration‐based measurement method due to the low shaft speed and light radial load condition specified for nondestructive inspection, which needs complicated and time‐consuming calculation and is not suitable for online inspection. Thus, there are problems about special operating conditions and weak fault severity in quality inspection of bearings, which is quite different from the fault diagnosis of bearings. In this paper, a novel dynamic quality evaluation technique is proposed based on the measurement of Hertz deformations. The measurement system is mainly composed of an eddy current sensor, sensor fixture, and data acquisition platform with less transfer path than the vibration‐based measurement system. The sensor fixture is optimized through numerical simulations to obtain signals with a high signal‐to‐noise ratio. Accurate evaluation of dynamic quality can be implemented reliably with simple signal processing. The proposed method can be used with a rotating speed of 100 rev/min and test load of 100 N, which is remarkably lower than the traditional quality inspection machineries with a rotating speed of around 1000 rev/min and the test load of 400 N. Both simulation and experiment studies have verified the proposed method.

Smart technology measurement for surface defects inspection in die casting

The objective of the paper is to demonstrate a measurement strategy for online inspection of surface defects in products, especially discontinuities which appear in castings after machining. The essence of the proposed online vision inspection system is to provide an automated method for obtaining and analyzing images of the inspected surfaces, to allow an unmistakable and consistent finding of defects and specifying their types. The proposed solution could improve productivity and quality in the manufacturing process especially for the automotive industry and significantly improve automatization in die casting technology. This is one of the most fundamental problems of a research studies in process design for automotive parts such as: handles of transmission systems, cylinder pistons and cylinder front faces in engine bodies, that has significant influence for the manufacturing cost. Proposed by the author an online vision inspection systems are responded to users demands for systems that are easier to use, low-cost, and flexible. Using a combination of a vision hardware and software, the proposed vision system could analyze the images, usually in a fraction of a second. Since the human visual inspection is slow and expensive, a smart technology is an alternative solution for the online inspection. The developed smart technology uses advance vision system with specially designed lighting, an advanced image processing algorithm for defect detection and their classification.

Self-powered visualization system by conjunction of photovoltaic effect and contact-electrification

Nanosensors play an irreplaceable role in environmental and health monitoring fields. Single-function sensor can’t meet the application requirements of increasingly complicated environment. In this work, a novel multifunctional self-powered visualization system (SVS) by photovoltaic effect and contact-electrification has been demonstrated. The fabricated SVS can detect ultraviolet (UV) light and mechanical signals without any external power，which can visualize and distinguish both signals accurately. The SVS shows excellent photoresponse properties as a self-powered UV photodetector, with a high sensitivity over 107, a rapid rise/decay time (less than 0.4 s), and a low detection UV light intensity of 0.24 mW/cm2. In addition, the SVS can detect gentle wind (1.0 m/s) as a self-powered wind velocity sensor system. It has been applied for the real-time environment monitoring of UV light and wind. The visualizing function of the SVS for different environmental signals has wide potential applications in human-electronic interfacing, Internet of Things, real-time online inspection and other intelligent fields.

Contactless high-speed eddy current inspection of unidirectional carbon fiber reinforced polymer

This paper presents the development and the results of a customized eddy current (EC) non-destructive testing (NDT) system for highly demanding online inspection conditions. Several planar eddy current array probes were designed, numerically simulated and experimentally compared for the inspection of low conductivity unidirectional carbon fibre reinforced polymer (CFRP) ropes. The inspections were performed using a dedicated scanner device at 4 m/s with 3 mm probe lift-off where defects under 1 mm were detected with an excellent SNR. Different defect morphologies and sizes, such as broken fibres and lateral cuts, were successful detected and compared to conventional probes.

FUZZY LOGIC CONTROL FOR VARIED INSPECTION

Varied inspection is an on-line dynamic inspection method in which the amount of inspection performed fluctuates, based on production needs. A two-level control architecture of fuzzy controllers was used to perform varied inspection. The research was the application of varied inspection into three manufacturing subsystems: transfer lines, assembly, and disassembly. The results from each simulation showed that varied inspection could be used to meet production cycle aims while reducing manufacturing lead time. Varied inspection was compared with 100 per cent inspection to determine the significance of varied inspection compared with traditional quality control methods.

A novel algorithm for tool wear online inspection based on machine vision

To inspect the tool wear condition in the process of numerical control (NC) machining for difficult-to-cut material, an online inspection system for tool wear is developed based on machine vision. With the help of MATLAB software, a self-matching algorithm is proposed according to the characteristics of tool wear images. The corresponding user-friendly graphical user interface (GUI) of the algorithm is developed. The bottom edges are separated by the adaptive connecting domain labeling to analyze wear condition of each edge. Then, each edge is arranged regularly by the improved rotatory positioning. The cutting edge is extracted by the method of partial angle threshold to fit and calculate the bottom wear value. It is shown that the absolute values of errors on the maximum wear width are less than 0.007 mm by using the self-matching algorithm. In the case of severe wear and breakage, the absolute values of errors on the maximum wear width are less than 0.057 mm because of uneven reflected light. The system features high response speed, high inspecting accuracy, and anti-noise performance. It is proved to be able to increase the utilization of cutting tools and guarantee the quality of workpiece. This method is potential to guarantee the reliability of cutting tool in aerospace manufacturing.

Tool path transplantation method for adaptive machining of large-sized and thin-walled free form surface parts based on error distribution

Abstract For the machining of freeform surface parts with large size and thin walls, the actual shape of the blanks is always inconsistent with the desired shape, unexpected overcut or undercut will be introduced by using the tool paths generated according to the nominal part model. According to the online inspection results of large-sized and thin-walled freeform surface part, we found that the error distribution of the blank has different magnitude and form over the whole surface of the part. In this paper, the error distribution is mathematically defined based on online inspection data, and new tool paths adaptive to the actual shape of the workpiece are generated on the basis of the tool paths of nominal model, different tool path transplantation strategies are adopted according to the error distribution for different surface areas. A typical aircraft skin part with dimensions of 3867 mm × 1905 mm × 6 mm is used as a case to validate the proposed approach, the actual cutting experiment results show that the wall thickness error ranges from –0.089 mm to + 0.162 mm, and the shape error is within 0.46 mm, which can both meet the machining requirement of the aircraft skin parts.

A precise magnetic flux leakage method for the defects detection within the steel thin sheet

Defects detection of ferromagnetic steel thin sheet is very important in metallurgical industry. In this article, a new magnetic flux leakage (MFL) technology is used to complete this task. In our work, a tiny permanent magnet is used to replace the conventional coils to magnetize testing specimen and precise Tunnel Magneto-resistive (TMR) sensor is employed because of its superior sensitivity and linear operation range over traditional magneto-resistive sensors. Two kinds of thin sheets were selected: galvanized sheet and cold-rolled sheet. The experimental results indicate that the present method is reliable for detecting defects of steel thin sheet, and defects with diameter of 400 and 20 μm can be detected clearly for galvanized sheet and cold-rolled sheet, respectively. This accuracy fully satisfies the metallurgical production requirement. This method exhibits its potential to realize in-situ, online inspection process of thin sheet production.

Design of fruits and vegetables online inspection system based on vision

This paper proposes a vision-based online fruit and vegetable inspection system. This system consists of two parts: the online detection of fruits and vegetables and the online weighing measurement of fruits and vegetables. Online detection system can detect the species of fruits and vegetables, and through the database to get the unit price of fruits and vegetables. The image processing includes image pre-processing, filtering, background removal, noise filtering and hole filling for recognition, and analysis of blocks. The weight of the fruits and vegetables is obtained by gravity sensor. Prices of fruits and vegetables can be displayed in real time in order to achieve the goal of unattended retail. This study built an online fruits and vegetables detection system which may be accelerated in the future for commercial uses.

Ultrasonic Excitation – Fiber Bragg Grating Sensing Technique for Damage Identification

Nondestructive testing plays an important role in on-line equipment inspection. Traditional nondestructive testing methods have been widely utilized for this. Ultrasonic excitation-fiber Bragg grating sensing technique is based on well-developed optical fiber grating technology, which has a good prospect of damage detection for mechanical equipments. However the corresponding research is still at the starting stage and further research work is necessary. The main contribution pursued in this investigation is to establish a detection system based on fiber Bragg grating sensing under ultrasonic excitation and predict the damage position through time delay on the plate structure. Differencing from the conventional approaches, a new way of damage detection utilizes fiber Bragg grating sensors and a position algorithm by two-dimensional method is derived. Moreover, wavelet transform is adopted in the subsequent signal processing. Finally, a platform for experiment is built to verify the theoretical analysis of the sensing characterization of the fiber Bragg grating sensor in ultrasonic excitation. Experiment results show that the wavelet transform is effective for signal denoising and our localization algorithm is feasible.