激光辐照下预加载CFRC层合板断裂行为实验研究

针对激光与机械载荷联合作用下碳纤维/环氧树脂增强复合材料（CFRC）层合板失效时间的预测需求,实验研究了不同激光功率密度（70～210 W/cm2）、不同预应力水平（拉伸强度的50%和70%）、不同光斑尺寸（拉伸试件宽度的70%和100%）下2 mm厚层合板的失效机理,获取了不同影响因素对断裂时间的影响规律。结果表明:预加载层合板失效机制为迎光面环氧树脂基底材料热解、纤维氧化断裂,背光面剩余结构偏脆性断裂;在预应力一定条件下,试件断裂时间与辐照激光功率密度成指数规律;预应力水平对断裂时间影响显著。     

木塑复合材料缺陷及损伤的声发射信号特征分析及神经网络模式识别

针对木塑复合材料五种典型的缺陷及损伤机制，选择合适的木塑试样，应用三点弯曲的加载方法采集声发射信号。对主损伤区附近的声发射事件，应用小波变换提取特征参数，确定五类主要损伤机制所对应的声发射信号特征。采用B—P型反向传播神经网络构成的智能化模式分类器，对此五类声发射信号进行识别，获得了满意的效果。     

基于中红外光声光谱的聚合物包膜控释肥料养分释放曲线预测

以聚丙烯酸酯类系列水基聚合物包膜控释肥料为样品,测定了包膜肥料养分的释放曲线并原位测定了肥料包膜的中红外光声光谱,分析了不同肥料的养分释放曲线以及不同包膜材料的红外光声光谱特征;采用广义回归神经网络模型(GRNN),以肥料包膜红外光声光谱的主成分作为GRNN模型的输入层,并以包膜肥料养分释放曲线为输出层,构建了预测养分释放曲线的GRNN模型。结果表明,GRNN模型能快速有效地预测包膜肥料养分释放曲线,其预测相关系数(R2)达0.93以上;包膜的探测深度明显影响释放曲线的预测误差,最小预测误差为7.14%,平均为10.28%,且基于包膜表层红外光声光谱的预测误差最小。因此,结合GRNN模型,红外光声光谱可为包膜肥料养分释放曲线的快速预测提供新手段。     

Acousto-ultrasonic sensing for delaminated GFRP composites using an embedded FBG sensor

In this paper, an active damage detection system for composite laminates is introduced. A fiber Bragg grating (FBG) was employed as an embedded sensor for detecting ultrasonic Lamb wave generated by a piezoelectric (PZT) actuator, inside the laminates. There were two different configurations for the system, making use of a broadband light source and a laser source, respectively. It was found that the use of a laser source could provide a higher sensitivity, so it was preferred in an ultrasonic signal acquisition unit for damage detection. When a delamination is developed in composite structures, the propagation characteristics of Lamb wave is then changed. According to altered signals, the types and natures of delamination inside the structures can be evaluated. In this experiment, glass fiber-reinforced epoxy (GFRP) composite beams were used to study the feasibility of our proposed detection technique. FBG sensors were embedded in different layers of the composite beams. The responses of FBG sensors from both intact and delaminated composite beams were then compared. Finally, the acquired Lamb wave signals corresponding to different delamination sizes and locations were examined.     

基于径向基函数神经网络的高光谱遥感图像分类

从径向基函数神经网络的理论出发,针对高光谱数据的特点,设计了有效的特征提取模型,再与径向基函数神经网络的输入层连接,建立了一个新的径向基函数神经网络的高光谱遥感影像分类模型,并用国产OMISII传感器获得的64波段数据进行试验。首先进行了最小噪声分离变换,提取了1～20个分量的数据,使用提取后的数据（20维）、提取后数据的纹理变换（20维）和主成分分析的前（20维）,组成了60维向量数据进行分类处理,这种分类器结构简单、容易训练、收敛速度快,其分类精度达到69.27%,高于BP神经网络分类算法(51.20%)以及常用的最小距离分类（MDC）算法(40.88%)。通过对结果和过程进行分析,实验证明径向基函数神经网络在高光谱遥感分类中具有较好的适用性。     

Properties of cementitious composites containing polypropylene fiber waste

The carpet waste fibers and waste glass that create a serious environmental problem can be converted into useful products. The use of these waste materials in a cement-based composite can be a promising direction for waste reduction and resources conservation. In this study, several tests have been carried out to investigate the performance of control mortar using recycled glass and fibers as a fraction of aggregates in a cement-based composite. These tests included compressive strength, flexural strength, flexural toughness and water absorption. The results revealed that carpet waste fiber and waste glass could be reused as substitutes for conventional materials in cement-based composites.     

Preparation,Characterization, and Properties of Resol-Based Benzoxazine Intermediates and Glass Cloth Reinforced Laminates Based on Their Polymers

A series of multifunctional benzoxazine intermediates (Boz-p-x) with different molecular weights and curing behaviors were synthesized based on resol. Their composition was controlled by adjusting the mole ratio (x) of aniline to phenol; i.e. the content of benzoxazine rings formed was varied. The chemical structures and hydrogen-bonding interaction of the intermediates were confirmed by ¹H nuclear magnetic resonance (NMR) spectroscopy and Fourier transform infrared (FTIR) spectroscopy. Their curing behaviors were investigated by differential scanning calorimetry (DSC). The results indicated that the residual phenolic hydroxyl groups in the intermediates formed two kinds of hydrogen bonding and acted as a catalyst in the curing process. The more residual phenolic hydroxyl present, the higher the glass transition temperature was and the faster the curing reaction. Further, glass cloth reinforced laminates based on poly(Boz-p-x)s were prepared. The mechanical properties of the laminates were tested by dynamic mechanical analysis (DMA) and flexural measurements. The poly(Boz-p-8) laminate exhibited the highest glass transition temperature (T _g) and mechanical properties in the laminates. Its T _g was 237°C and its flexural modulus and flexural strength were 26.8 GPa and 597 MPa, respectively. The laminates of poly(Boz-p-6), poly(Boz-p-8), and poly(Boz-p-10) showed good modulus retention at elevated temperature due to more ?OH???N hydrogen-bonding interactions. In addition, poly(Boz-p-x)s exhibited high thermal stability and char yield at 800°C.     

Fabrication and Properties of Recycled Cellulose Fibre-Reinforced Epoxy Composites

Epoxy matrix composites reinforced with recycled cellulose fibre (RCF) were fabricated and characterized with respect to their flexural and impact properties. Reinforcement of the epoxy by RCF resulted in a significant increase in the strain at failure, fracture toughness and impact toughness but only a moderate increase in flexural strength and flexural modulus. The effect of accelerated exposure to seawater on the flexural and impact properties was also investigated. The salient toughening mechanisms and crack-tip failure processes were identified and discussed in light of observed microstructures, in particular the orientation of RCF sheets to the applied load.     

Ply-by-Ply Failure Analysis of Laminates Under Dynamic Loading

Ply-by-ply failure analysis of symmetric and anti-symmetric laminates under uniform sinusoidal transverse dynamic loading is performed for a specified duration. The study investigates the first ply failure load, followed by the detection of successive ply failures along with their failure modes using various failure theories. Some of the well-established failure theories, mostly used by the researchers, are considered for the failure prediction in laminates. The finite element computational model based on higher order shear deformation displacement field is used for the failure analysis and the complete methodology is computer coded using FORTRAN. The ply-discount stiffness reduction scheme is employed to modify the material properties of the failed lamina. The failure theories used in the analysis are compared according to their ability to predict failure load, failed ply, failure mode and progression of failure. The failure analysis is performed for both the cross-ply and angle-ply laminates with all edges simply supported and clamped. The significance of fibre orientation and stacking sequence in terms of the strength of a laminate and failure progression is also highlighted.     

Mechanical behavior and health monitoring by Acoustic Emission of unidirectional and cross-ply laminates integrated by piezoelectric implant

Recent progress in sensor technologies, signal processing and electronics has made it possible to fulfill the need for the development of in-service structural health monitoring (SHM) systems. This study presents a health monitoring of composite materials integrated by piezoelectric sensor using Acoustic Emission (AE) technique. A series of specimens of composite laminates with and without piezoelectric implant were subject to three-point bending in static and creep tests while continuously monitoring the response by the AE technique. The analysis and observation of AE signals lead to the identification of the acoustic signatures of damage mechanisms in composite laminates. The mechanical behavior of composites with and without integrated sensor shows no difference in the form. The incorporation of piezoelectric sensor influences specially the fracture load and causes low degradation of mechanical properties of materials. One of the major differences between the two types of materials (with and without embedded sensor) is the intense acoustic activity in the integrated material.     

Preparation and characterization of carbonyl iron/glass composite absorber as matched load for isolator

Composite absorbers made from 66 wt% carbonyl iron and 34 wt% low melting point glass powder were prepared by a pressureless sintering technique in a nitrogen atmosphere. Apparent porosity and bending strength of the as-prepared composites were investigated. The microstructure, heat resisting properties and electromagnetic properties were characterized by scanning electron microscopy, thermal gravimetric analysis–differential scanning calorimetry and vector network analyzer. The results show that the carbonyl iron/glass composite absorbers were difficult to densify. As the sintering temperature and soaking time increased, the apparent porosity first decreased and then increased, whereas the bending strength showed the opposite change. The composite absorber sintered at 520 °C for 40 min achieved the minimum apparent porosity of 13.08% and the highest bending strength of 52 MPa. Compared to the carbonyl iron/silicone rubber absorber, the carbonyl iron/glass composite absorber exhibited better heat resisting properties, and the initial oxidation temperature was increased about 200 °C. The composite absorber with a thickness of 1.25 mm showed a good microwave absorbing property in 8–12 GHz.     

Experimental analysis of the linear and nonlinear behaviour of composites with delaminations

An analysis of the linear and nonlinear vibration responses of composites with delaminations is presented. The effect of delamination size on the linear and nonlinear vibration response is studied. The composite material used in this paper is a glass fibre reinforced plastic (GFRP) having delaminations at the plies interfaces. The experimental procedure consists in inducing the specimen on its resonance flexural modes with different excitation levels (amplitudes) for six bending modes and for each delamination length. The presence of the nonlinearity introduced by the delamination was clearly identified by the variation of natural frequencies for increasing excitation levels. Then, nonlinear elastic parameters for progressive delamination length were determined and discussed for the first six bending modes. The linear and the nonlinear elastic parameters were compared in their sensitive modes.     

Correlation of the mechanical properties to the fiber–matrix adhesion of Nextel 312™ fiber/BN/Blackglas™ composites

Boron nitride (BN)-coated aluminoborosilicate (Nextel? 312) fibers, produced via ammonia nitridation, along with 'as-received' and 'desized' fibers, were composited in a silicon oxycarbide (Blackglas?) matrix. The mechanical properties, failure properties, and fiber–matrix interfacial chemistry of the composite were investigated. BN treated fiber composites show a 90% improvement in flexural strength and substantial increases in shear strength (short beam shear and Iosipescu) over the 'as-received' fiber composite. The composite fabricated with 'desized' fibers underwent spontaneous delamination during pyrolization, precluding mechanical testing. X-ray photoelectron spectroscopy of the starting materials and of composite fracture surfaces combined with scanning electron microscopy and energy dispersive X-ray spectroscopy indicate that the locus of failure of the BN-coated fiber composite occurs at the matrix/BN coating interface.     

小波变换近红外光谱结合径向基神经网络快速分析异福片

应用小波变换(WT)处理近红外漫反射光谱结合径向基神经网络(RBFNN)建立快速分析异福片中利福平和异烟肼含量的模型(WT-RBFNN)。用小波变换的低频系数作为RBFNN的输入节点, 研究了网络参 数包括隐含层神经元数和径向基宽度(SC)对模型的影响。与经典的RBFNN和PLS相比较表明, WT-RBFNN模型压缩了原始光谱, 除去了噪音和背景的影响, 拟合效果很好。优选的WT-RBFNN模型对校正集样品 中利福平、异烟肼的交互验证均方根误差(RMSECV)分别为0.006 04和0.004 57;对预测集样品预测均方根误差(RMSEP)分别为0.006 39和0.005 87。同时预测集样品中利福平和异烟肼与RP-HPLC测定结果 的回归系数(r)分别为0.995 22和0.993 92, 相对误差在2.300%以下。这些结果显示了该方法建模的稳健性和模型的预测精度均很高, 同时此方法具有非破坏、无污染、可在线检测等优点, 对替代常规药物 分析方法有重要的意义。     

Generalized unscented Kalman filtering based radial basis function neural network for the prediction of ground radioactivity time series with missing data

On the assumption that random interruptions in the observation process are modeled by a sequence of independent Bernoulli random variables, we firstly generalize two kinds of nonlinear filtering methods with random interruption failures in the observation based on the extended Kalman filtering (EKF) and the unscented Kalman filtering (UKF), which were shortened as GEKF and GUKF in this paper, respectively. Then the nonlinear filtering model is established by using the radial basis function neural network (RBFNN) prototypes and the network weights as state equation and the output of RBFNN to present the observation equation. Finally, we take the filtering problem under missing observed data as a special case of nonlinear filtering with random intermittent failures by setting each missing data to be zero without needing to pre-estimate the missing data, and use the GEKF-based RBFNN and the GUKF-based RBFNN to predict the ground radioactivity time series with missing data. Experimental results demonstrate that the prediction results of GUKF-based RBFNN accord well with the real ground radioactivity time series while the prediction results of GEKF-based RBFNN are divergent.     

Fiber optic acoustic emission sensor and its applications in the structural health monitoring of CFRP materials

A fiber optic acoustic emission sensor based on fused-tapered coupler and its applications in structural health monitoring are proposed in this paper. The fiber optic acoustic emission sensor (FOAES) was tested using pencil lead break tests compared with the commercial acoustic emission sensor (R15 PZT). Besides, the sensor was embedded into the carbon fiber reinforced plastics (CFRP) materials and tested in the same way. It successfully detected the AE signals. FOAES was applied in the structural health monitoring (SHM) of CFRP materials. Damages of carbon fiber/epoxy composite laminates during three-point-bending test were monitored by surface-mounted and embedded FOAES, respectively. Results identified that the sensor embedded into composite structures could monitor damage of composite laminates on-line as the surface-mounted sensor did.     

人工神经网络结合近红外光谱用于木材树种识别

测量了不同产地及品种的89个木材样品的近红外光谱,并分别使用反向传播人工神经网络(back propagation artificial neural networks,BPANN)与广义回归神经网络(generalized regression neural network,GRNN)建立了NIRS树种识别模型。通过方差分析分别选择两种神经网络所用参数,并采用最优参数进行网络训练。考虑到样品光谱的差异,对含不同水平白噪声与不同水平偏置的光谱进行模拟,并使用建立的模型对模拟光谱进行预测。发现两种神经网络模型均有较好的预测结果,其中BPANN模型,对含偏置水平不高于2％、噪声水平不高于4％的模拟光谱识别正确率在97％以上;GRNN模型,对含偏置水平不高于2％、噪声水平不高于4％的模拟光谱识别正确率在99％以上。     

Role of curing conditions and silanization of glass fibers on carbon nanotubes (CNTs) reinforced glass fiber epoxy composites

Curing behavior of amino-functionalized carbon nanotubes (ACNT) used as reinforcing agent in epoxy resin has been examined by thermal analysis. Experiments performed as per supplier’s curing conditions showed that modification of the curing schedule influences the thermo-mechanical properties of the nanocomposites. Specifically, the glass transition temperature (T_g) of ACNT-reinforced composites increased likely due to the immobility of polymer molecules, held strongly by amino carbon nanotubes. Further, a set of composites were prepared by implementing the experimentally determined optimal curing schedule to examine its effect on the mechanical properties of different GFRP compositions, while focusing primarily on reinforced ACNT and pristine nanotube (PCNT) matrix with silane-treated glass fibers. From the silane treatment of glass fibers in ACNT matrix composition it has been observed that amino silane is much better amongst all the mechanical (tensile and flexural) properties studied. This is because of strong interface between amino silane-treated glass fibers and modified epoxy resin containing uniformly dispersed amino-CNTs. On the other hand, PCNT GFRP composites with epoxy silanes demonstrated enhanced results for the mechanical properties under investigation which may be attributed to the presence of strong covalent bonding between epoxy silane of glass fiber and epoxy–amine matrix.     

碳纤维-泡沫铝夹芯板低速冲击响应

为研究夹芯结构的低速冲击响应,以碳纤维(T700)/环氧树脂复合材料层合板为上下面板,以闭孔泡沫铝为芯层,模拟夹芯板落锤冲击时的损伤演化过程。复合材料层合板采用三维实体单元建模,基于有限元软件ABAQUS中的用户子程序VUMAT,引入三维Hashin失效准则模拟复合材料的损伤破坏;采用二次应力准则,Cohesive单元模拟黏结层的层间失效;闭孔泡沫铝芯层采用3D Voronoi细观模型建模。分析复合材料夹芯结构在落锤冲击下的损伤起始、损伤扩展和最终破坏模式,通过锤头的接触力、位移、夹芯板的内能、后面板的最大位移研究夹层结构的能量吸收情况及抗冲击特性,得出了在质量保持不变的情况下,5种芯层相对密度和厚度的耦合关系中的最优设计是芯层相对密度15.0%,厚度为10 mm,为满足实际工程中的需求提供了设计依据。     

Application of quantitative structure-activity relationship to the determination of binding constant based on fluorescence quenching

Quantitative structure-activity relationship (QSAR) model was used to predict and explain binding constant (log K) determined by fluorescence quenching. This method allowed us to predict binding constants of a variety of compounds with human serum albumin (HSA) based on their structures alone. Stepwise multiple linear regression (MLR) and nonlinear radial basis function neural network (RBFNN) were performed to build the models. The statistical parameters provided by the MLR model (R²=0.8521, RMS=0.2678) indicated satisfactory stability and predictive ability while the RBFNN predictive ability is somewhat superior (R²=0.9245, RMS=0.1736). The proposed models were used to predict the binding constants of two bioactive components in traditional Chinese medicines (isoimperatorin and chrysophanol) whose experimental results were obtained in our laboratory and the predicted results were in good agreement with the experimental results. This QSAR approach can contribute to a better understanding of structural factors of the compounds responsible for drug-protein interactions, and can be useful in predicting the binding constants of other compounds.