Prepreg tack is an important process quality parameter for prepregs during laying. Aiming at the current lack of standardized testing for prepreg tack, this paper established a quantitative testing method for prepreg tack—a compression-to-tension method—and proposed a parameter of Compression Tack Index as a quantitative evaluation index for prepreg tack. The prepreg/prepreg tack and prepreg/metal tack of carbon fiber reinforced epoxy prepregs were evaluated and the applicability of this compression-to-tension method was verified, comparing it with the qualitative testing method by vertical metal plates. The results show that the compression-to-tension method is suitable for quantitative testing of the tack for unidirectional prepregs and fabric prepregs, with good repeatability and stability of test results, and is not affected by personnel changes. Considering that tack characterization based only on the separation process cannot accurately evaluate the tack of different materials, Compression Tack Index is an accurate parameter that characterizes the prepreg tack because it can reflect the process of tack formation and tack separation. Compared with the vertical metal plate method, the discrimination of the test results by the compression-to-tension method is significant. The tack of the slitting prepreg without polyethylene film coating is lower than that of the mother prepreg (one-meter-width prepreg) with polyethylene film.

Fuzz defects are prevalent surface imperfections in carbon fiber-reinforced polymer (CFRP) prepregs. Current manual inspection methods or conventional neural network-based approaches face significant limitations in achieving standardized and accurate severity assessment of such defects. In this article, a methodology comprising three key technical innovations is proposed: First, an adaptive thresholding algorithm is implemented, utilizing local average grayscale values to accurately identify fuzz defect pixels. Second, a grayscale histogram analysis is performed on the identified defect regions to quantify severity levels, effectively mitigating the influence of substrate material variations and illumination conditions on assessment accuracy. Third, a quantitative formula is defined based on the detection boxes drawn by the neural network object detection model and the effective area of defects to evaluate the severity of fuzz defects. Experimental validation shows 90% consistency with practical manual assessment in defect severity ranking tasks, proving its industrial applicability.

Aim to the current lack of standardized testing for prepreg tack, this paper established a quantitative testing method for prepreg tack - a compression-to-tension method, and proposed a parameter of Compression Tack Index (CTI) as a quantitative evaluation index for prepreg tack. The prepreg / prepreg tack and prepreg / metal tack of carbon fiber reinforced epoxy prepreg were evaluated and verified the applicability of this compression-to-tension method, comparing with the qualitative testing method for vertical metal plates. The results show that the compression-to-tension method is suitable for quantitative testing of the tack for unidirectional prepregs and fabric prepregs with a good repeatability and stability of test results, and is not affected by personnel changes. Compared with the vertical metal plate method, the discrimination of the test results by the compression-to-tension method is significant. The tack of the slitting prepreg without PE film coating is lower than that of the mother prepreg with PE film.

Composite T-stiffened skins, widely used in aeronautical structures nowadays, are fabricated by co-curing and co-bonding processes in this work. An experimental study was focused on the interface fracture progression of T-stiffened skins under tensile loading, including the macroscopic failure process and the microscopic failure mode at several bonding interfaces. The experimental results showed that a similar damage evolution process is found in co-curing and co-bonding processes. Damage is found to initiate in the triangle filling area or the stiffener/filler bonding interface, and it gradually extends to the web part, and finally rapidly extends to the stiffener/skin bonding interface with a complete separation failure. The final failure position occurs in the skin layers closest to the adhesive layer under the stiffener in the co-bonding process with a failure load of 39% higher than the initial damage load, while the final failure position occurs at the flange/skin bonding interface in the co-curing process. The mixed crack growth of mode I and mode II is the main failure mode at the bonding interfaces of composite T-stiffened skins.

A comprehensive consideration of delamination location, delamination size and delamination quantity was focused to determine a quantitative evaluation of the influence of delamination on the tensile strength of T-stiffened panels under the out-of-plane tensile load. Moreover, the failure behavior of T-stiffened panels without delamination and those with delamination under tensile loading was also examined. A characteristic parameter about delamination in the stiffener/panel interface was proposed by means of the ratio (S-r) of the delamination size in the filler/panel interface to the filling area in the filler/panel interface. It is found that the delamination located closely to the stiffener/panel interface, especially closely to the filler/panel interface has the most significant influence on the tensile strength. Another interesting observation is that delamination in the stiffener/panel interface and the stiffener corner also changes the failure mode which cracks along the interior of panels near the adhesive layer.

分层是复合材料层合板最主要的缺陷/损伤形式,当前研究多集中于复合材料使用过程中的分层损伤,对制造过程中的分层缺陷研究较少.复合材料层合板的分层缺陷在整体成型过程中因不均匀温度场、非对称结构等原因易发生扩展,采用预埋隔离纸模拟分层缺陷,研究分层缺陷在整体成型过程中的扩展行为,对分层扩展的驱动力进行了计算分析,并通过分析热循环对T300/QY8911复合材料层合板界面性能的影响以及计算裂纹尖端能量释放率,从实验研究和有限元计算两方面证明了分层扩展是一个动态过程.研究结果表明,热残余应力是T300/QY8911复合材料层合板整体成型过程中发生分层缺陷扩展的主要驱动力;增加热循环次数和提高热循环温度会显著降低T300/QY8911复合材料层合板的层间剪切强度,当能量释放率降低到低于层间断裂韧性值时,分层动态扩展过程则停止.

针对复合材料构件热压罐成型过程中常见的分层缺陷,考察了整体成型工艺温度对分层扩展、QY8911双马树脂基体韧性及T300/QY8911层合板Ⅰ型层间断裂韧性的影响,并通过分层扩展断面形貌深入分析了复合材料整体成型工艺中分层扩展的路径和断面破坏模式,给出了复合材料整体成型工艺和结构设计的优化建议措施.结果表明,随着整体成型最高温度的升高,分层扩展程度增大,QY8911双马树脂基体的拉伸强度和拉伸模量逐渐降低,T300/QY8911层合板Ⅰ型层间断裂韧性逐渐增大;对分层扩展断面进行SEM扫描电镜分析发现分层扩展沿着层间开裂,断面内存在基体断裂和基体/纤维界面脱粘两种破坏模式,Ⅰ型层间断裂是复合材料整体成型工艺中分层扩展的典型微观特征.

For composite T-stiffened skins widely used in the aircraft industry nowadays, the triangle filling area of this structure is the weakest part of compaction quality and mechanical properties. In this paper, T-stiffened skins were prepared by the integral molding process of co-bonding, with a focus on the influence of the tool assembly schemes and different curvature radii on the fiber compaction and tensile strength of the triangle filling area. Three quantitative characteristic parameters of the fiber compaction in the triangle filling area were proposed: stiffener thickness non-uniformity (H-d), fiber volume fraction (V-f) and deformation index (a). The results show that an uneven fiber distribution is found in the inner of the triangle filling area, and the maximum difference of V-f is almost 10%. The deformation index of the triangle filling area is the closest to the theoretical value when rigid tool with triangle flexible tool assembly scheme is adopted. An increase of curvature radius (especially greater than 5 mm) causes a decrease of the stiffener thickness non-uniformity, a decrease of the deformation index of the triangle filling area which is close to the theoretical value of 0.5 for the curvature radius of 7 mm and 9 mm, and an increase of the tensile strength (initial damage load and failure load) of T-stiffened skins under the out of plane tensile load with the highest tensile strength for the curvature radius of 7 mm.

分层缺陷是复合材料构件热压罐成型过程中常见的一种缺陷形式,其种类繁多、成因复杂,严重影响了复合材料构件使用的可靠性.本文在归纳分层缺陷种类及成因分析的基础上,系统开展了分层缺陷实验模拟方法研究,并按照复合材料构件整体成型工艺温度制度考核了四种预埋分层缺陷的扩展情况.研究结果表明,埋入防粘纸产生的分层与实际的制造分层最为类似;埋入聚四氟乙烯薄膜与真空袋与上下两层的铺层结合成为一体形成夹杂缺陷,而不是分层;由于聚四氟乙烯布上有孔,也不宜用来模拟分层;同时,防粘纸最易引起分层扩展,对整体成型工艺质量的影响最大.

Composite T-stiffened skins widely used in aircraft industry nowadays were fabricated by using the co-curing, co-bonding and secondary bonding processes. The stiffener thickness and the tensile property of T-stiffened skins under tensile loading were performed as evaluation indexes to evaluate the manufacturing quality and mechanical property of T-stiffened skins, respectively. The results show that the better manufacturing quality and higher tensile property of T-stiffened skins are found in the co-bonding process, and the tensile property increases with the increase of the adhesive film thickness in the co-bonding process. The co-curing process provides a similar tensile property of the co-bonding process with 0.15 mm thickness adhesive film, but different failure modes in the stiffener/skin interface are found in co-curing and co-bonding processes. Moreover, the filler content and the filler type in the triangle filling area of T-stiffened skins also have significant influence on the tensile property of T-stiffened skins, and rolling adhesive films as fillers into the triangle filling area exhibits the highest tensile property in five kinds of fillers during the co-bonding process.

Defect information obtained by non-destructive identification for more than four thousand composite components fabricated by convex mould in autoclave molding was statistically analyzed in this paper. The correlated rules between curvature radius of composite component and manufacturing defect were revealed. By self-designed solid pressure distribution testing method, the pressure distribution in the interface between flexible mould and rigid convex mould was also measured. Furthermore, combined with stress analysis in corner section and defect micrograph, the forming mechanism of the defect was primarily analyzed. The process control measures to avoid manufacturing defect were given. The results indicate that the defect ratio and the size of delamination in corner section of composite component both decrease with the increase of the curvature radius of component. The thickness distribution tends to be thick in the corner and thin on both sides of flat part, which is related to the coupling action between percolation mechanism and shear flow mechanism and the uniformity of pressure distribution.

With the rapid increase in social websites that has dramatically increased the volume of social media, which includes the use of images and videos, visual understanding has attracted great interest in several areas such as multimedia, computer vision, and pattern recognition. Valuable auxiliary resources available on social websites, such as user-provided tags, aid in the tasks of visual understanding. Therefore, several methods have been proposed for exploring the auxiliary resources for tag refinement, image retrieval, and media summarization. This work conducts a comprehensive survey of recent advances in visual understanding by mining social media in order to discuss their merits and limitations. We then analyze the difficulties and challenges of visual understanding followed by several possible future research directions.

Recent years have witnessed the explosion of multimedia news and currently we are living in a sea of multimedia news. Faced with so huge quantities of multimedia news from all corners of the world, people are often confused and feeling overwhelmed because it takes much time and effort for them to access their desired news. As a consequence, multimedia question answering (QA) with the goal of providing precise answers to the query questions has attracted increasing attention in recent years. Towards this end, this work proposes solutions for multimedia news QA to provide an effective news information browsing experience by integrating multimedia information from multiple sources. Specifically, we first collect a large-scale multimedia news data set for question answering. And then topic analysis is used to present concise and reliable answers including textual and visual information. Experiments on the news dataset collected from multiple news websites demonstrate the encouraging performance of the proposed solutions.

Nowadays with the growing popularity of vehicles, traffic accidents occur more frequently, causing lots of casualties. In this paper, in order to avoid the accident where a vehicle collides with the one ahead, we present a novel vehicle brake behavior detection method by using a colorful camera or mobile device fixed on the windshield of the test car which utilized to capture the front vehicle information. The brake behavior detection in our work includes two procedures, brake lights region detection and brake behavior decision. For the first procedure, we use threshold segmentation and proposed horizontal–vertical peak intersection strategy to filter and generate the credible rear-light regions of the front vehicle in the YCrCb color space converted from the original RGB color space. For the second procedure, the sophisticated SVM classifier is trained to detect the brake behavior of the front vehicle. In this procedure, we extract discriminative features of the rear-light regions generated from the first procedure and then the features are used as the input of the pre-trained classifier. Extensive experiments on various real-world vehicle datasets demonstrate the effectiveness and real-time performance of our proposed brake detection strategy.

It is a necessary but challenging task to relieve users from the proliferative news information and allow them to quickly and comprehensively master the information of the whats and hows that are happening in the world every day. In this article, we develop a novel approach of multimedia news summarization for searching results on the Internet, which uncovers the underlying topics among query-related news information and threads the news events within each topic to generate a query-related brief overview. First, the hierarchical latent Dirichlet allocation (hLDA) model is introduced to discover the hierarchical topic structure from query-related news documents, and a new approach based on the weighted aggregation and max pooling is proposed to identify one representative news article for each topic. One representative image is also selected to visualize each topic as a complement to the text information. Given the representative documents selected for each topic, a time-bias maximum spanning tree (MST) algorithm is proposed to thread them into a coherent and compact summary of their parent topic. Finally, we design a friendly interface to present users with the hierarchical summarization of their required news information. Extensive experiments conducted on a large-scale news dataset collected from multiple news Web sites demonstrate the encouraging performance of the proposed solution for news summarization in news retrieval.

A field experiment was conducted to reveal the effects of nitrogen rate on super rice grain yield and its mechanism with two super rice cultivars,Liangyoupeijiu (two-line indica hybrid)and Huaidao 9 (japonica),and two high-yielding elite check cultivars,Shanyou 63 (indica hybrid)and Yangfujing 8 (japonica)as material at four nitrogen (N)rates,no N application and N rates of 120 kg/hm2 ,240 kg/hm2 ,and 360 kg/hm2 .The results showed that the grain yield of two super rice cultivars increased with increasing N rates from 0 to 360 kg/hm2 .Similarly,the physiological performance,such as net photosynthetic rate and nitrate reductase activity of the flag leaf,root oxidation activity and zeatin + zeatin riboside content,shoot dry weight,root dry weight and the root/shoot ratio were improved with increassing N rates for both super rice cultivars.The grain yield and the physiological performance of the two check cultivars were also improved with the increase of N rates from 0 to 240 kg/hm2 .The above-mentioned indexes were decreased at the N rate of 360kg/hm2 for the two check cultivars.At the same N rate the grain yield of the two super rice cultivars was higher than that of the two check cultivars,while the physiological performance of the super rice cultivars was just the opposite under N rates no more than 240kg/hm2 .However,the physiological performance was better for super rice at the N rate of 360 kg/hm2 .The results suggested that super rice cultivars could obtain a higher grain yield than elite high-yielding cultivars at either a low N rate or a high N rate,and a high N rate was favorable to realize high-yielding potential of super rice.

Application of nitrogen (N) during the mid-season is an important practice in rice production. However, little is known if its timing should vary with the varieties differing in panicle size. We investigated the effectiveness of various mid-season N applications on the grain yield and nitrogen use efficiency (NUE) for rice varieties differing in panicle size. Three high-yielding japonica rice varieties, Yangfujing-8, Lianjing-7 and Huaidao-9 with a small, medium and large panicle size respectively, were grown in the field. Four N treatments during the mid-season, applied at the stages of panicle initiation (PI), spikelet differentiation (SD), heading time (HT) or all the three stages, were adopted with no N application as control. Results show that their effectiveness on grain yield and NUE varied substantially with the variety panicle size. For Yangfujing-8, N applied at PI was the most effective in increasing yield and NUE, while for Huaidao-9, N applied at SD or HT was the most effective. The effectiveness was similar for Lianjing-7 when N applied at PI, SD or N split application at PI, SD, and HT. Nitrogen applied at SD markedly increased crop growth rate during the 15-day period proceeding heading and pre-anthesis non-structural carbohydrate accumulation in the stem at heading. While N applied at HT significantly increased the photosynthesis of leaves. Both treatments of N applied at SD and HT could increase activities of sucrose synthase and adenosine diphosphoglucose pyrophosphorylase in the inferior spikelets of HD-9 at the mid and late grain filling stages. According the results, a mid-season N application strategy is suggested: applying N at PI to increase sink capacity for a variety with a small panicle size, at SD or HT to increase filling efficiency for a variety with a large panicle size, and at either PI or SD for a variety with a medium panicle size. (C) 2013 Elsevier B.V. All rights reserved.

Today social media websites have made the great success, which allow users to upload personal media data and encourage them to annotate media data with free tags. With the rich tags, users can more conveniently access the image content on these websites. However, the tags annotated by users are generally in a random order without recovering any important relevant information to the content of images. To target this issue, we propose a tags ranking scheme to automatically rank the tags with respect to given images by taking both their relevance to this image's visual content and their relationships into account. The proposed framework mainly comprises four stages. First, given a tag query, a set of web images is collected from multiple searching engines to cover the semantic space. Second, initial relevance scores of the tags with respect to a given image's visual content are estimated in Bayesian framework, in which a fused visual similarity is adopted through the linear combination of global and local visual similarities. Third, tags graph is constructed by mining the relationships among tags. At last, a random walk process over tags graph is performed to refine the tags ranking. Experimental results showed that the proposed method is both effective and efficient.

As smart phone features dramatical performance, including faster processor, GPS chip, and more advanced camera, etc., more and more people use it for photography. With more photos stored on personal devices, there is a growing need of photo tools that can empower users to organize and manage personal photographs. To target this goal, we devote our efforts to develop a photos management system used for cellphone, namely Photo 4W, offering the index of mobile photography related to what, where, who and when. The key technologies in this system mainly include face annotation and scene classification. Face annotation is implemented by using interactive method after face detection. To help user annotate faces, we provided a batch solution in which users are allowed to multi-select a group of faces and then assign one name from a predicted candidate list. Scene classification utilizes semi-automated method to classify photos based on their semantic content. Experiments conducted on a real family album demonstrate that the proposed approach is effective and efficient for mobile photo management.