Journal of Aerospace Engineering

Owing to current limitations on seismic forecasting, there is a high chance that a considerable number of vehicles will remain on bridges when an earthquake occurs. In traditional seismic analyses, traffic loads were often ignored. Mode-based bridge–traffic interaction analysis usually cannot consider the nonlinearity effects of bridges under earthquakes, which are critical to short- and medium-span bridges. Traditional nonlinear seismic analyses using commercial or open-source software cannot directly incorporate complex dynamic interactions between moving vehicles and bridges. There is no reported methodology so far that can be used for nonlinear seismic analyses of typical short- and medium-span bridges while rationally considering the coupling effects between the bridge, moving vehicles, and an earthquake at the same time. A hybrid simulation approach is proposed to conduct nonlinear seismic analyses of bridge–traffic–earthquake systems by integrating stochastic traffic flow simulation, a mode-based fully coupled simulation technique of bridge–traffic systems, and a nonlinear seismic analysis platform developed based on OpenSees. A skewed and curved bridge, which is a design commonly used to overcome complex intersections and terrain restrictions for short- and medium-span bridges, is studied as a demonstration followed by the numerical investigation of the bridge’s seismic performance and the impact of incorporating traffic loads. The results suggest that the proposed hybrid methodology can capture complex dynamic interactions between bridges and multiple vehicles as well as nonlinear seismic performance to provide rational prediction results.

In the present study, a composite cantilever beam subjected to shear loading has been analyzed. A simplified procedure is presented to examine the composite cantilever beam. The governing differential equations and boundary conditions are established by applying the principle of minimum potential energy, and solutions to the differential equation are given. A very simple and convenient formula to calculate the bending stresses consisting of shear lag in a composite cantilever beam is derived, which has a similar form as that of the bending stress in the elementary beam theory (EBT). A numerical example is illustrated to demonstrate the simplicity and accuracy of the proposed simplified method. For $E_s{I}_s/E_o{I}_o=0.731$, the stress factor (${\sigma }_x/\sigma$), i.e., the ratio of actual flange stress to the stress calculated by EBT in the central line of the cover sheet at the clamped end, is computed as the following: 1.120, corresponding to the uniformly distributed load; 1.067, corresponding to the point load; 1.165, corresponding to the uniformly varied load increasing toward the support and 1.102 corresponding to the uniformly varied load decreasing toward the support. The results obtained by the simplified method have been verified by finite-element analysis (FEA). Further, the present methodology is compared with the Reissner box beam methodology. The theoretical results are found to compare well with test results and literature.

Using the experimental data of a wind-induced pressure coefficient, equations for the group method of data handling neural network (GMDH-NN) are developed to predict surface mean pressure coefficients ($\overline{C_p}$) on the frontal surface of different C-shaped building models. Toward this objective, an extensive experiment was carried out to obtain pressure coefficients over the surfaces of the models with varying configurations, corner curvatures, and angles of incidence in a subsonic wind tunnel. The input variables include the curvature ratio ($R/D$), overall side ratio ($D/B$), side ratio without curvature ($d/b$), height ratio ($D/H$), and angle of incidence ($\theta$) in the radian in the GMDH-NN to develop the model equation. The performance of the GMDH-NN equation is compared with two different methods, namely, the nonlinear regression (NLR) approach through a gene expression programming (GEP) technique and a feed forward neural network (FFNN) through different statistical measures. The results indicate that the proposed GMDH-NN equation satisfactorily predicts the $\overline{C_p}$ on the frontal surface with coefficients of determination ($R^2$) as 0.989 and 0.985 and the scatter index ($SI$) as 0.10 and 0.11 for training and testing data, respectively.

A wind tunnel experiment was conducted to measure the flow over a suction side of a DAE51 airfoil at low Reynolds numbers using particle image velocimetry (PIV). Five chord Reynolds numbers ranging from 39,000 to 118,000 were investigated at angles of attack ranging from 0° to 10° at an interval of 1°. The locations of laminar separation, transition onset, transition end, and reattachment were identified by examining the measured averaged and RMS velocity data. The locations determined from the PIV data were in very good agreement with the features of measured surface pressure distribution. A large amplification of the RMS velocity leading to transition of a separated shear layer was observed centered at the inflection point of the averaged velocity profile, which indicates the important role of inflectional instability on the transition onset. The behaviors of those locations with respect to the Reynolds number and angle of attack were analyzed. For a fixed Reynolds number, the size of the separation bubble decreases quickly, and the bubble moves upstream as the angle of attack increases. The transition onset location moves upstream with increasing Reynolds number and angle of attack. The transition length becomes shorter as the Reynolds number increases. The lowest angle of attack at which a separation bubble is formed on the airfoil surface depends significantly on the Reynolds number. The movement of the separation point with respect to the Reynolds number indicates a different tendency depending on the presence of the separation bubble.

Stay cables are prone to parametric vibrations, which might severely affect the serviceability and safety of cable-stayed bridges. This paper investigates the stability and dynamic characteristics of in-plane parametric vibrations for stay cables in cable-stayed bridges with superlong span under axial excitations. First, the nonlinear motion equation of a stay cable under axial harmonic excitation is derived on the basis of an established cable sag curve equation with the chord component of cable self-weight considered. Next, the equation is solved theoretically using a multiscale method, which reveals the conditions for the existence of the constant solution in parametric resonances. Subsequently, the sufficient conditions for the existence of the asymptotic steady solution are further obtained according to the Lyapunov’s first approximate stability criterion. Moreover, a numerical analysis is performed to assess the influences of the cable’s damping ratio, excitation frequency, and amplitude on the cable vibration using the longest cable, S36, in the Hutong Yangtze River Railway Bridge. The results show that the minimum external excitation amplitude exists for parametrical vibrations of stay cables, while the damping does not effectively suppress the amplitude of large-scale parametrical vibrations. Therefore, controlling the excitation amplitude may be an effective way to inhibit the parametrical vibrations of stay cables.

To support the design of a two-dimensional turbine blade profile with reduced shock loss, an efficient model for shock system with separation shock in highly loaded turbine cascade needed to be established to realize parameterized shock losses prediction and analysis. The discussions on the model demonstrate that the effect of base pressure on shock losses can be achieved by controlling the width of the base region (the region between the pressure surface and suction surface trailing edge separation points). Increasing the trailing-edge wedge angle can increase the base pressure as well as the width of the base region, thus reducing the shock losses. Moreover, the pressure change caused by reducing the suction surface curvature downstream of the throat can reduce the loss caused by the reflected shock.

This paper presents a hybrid model-based fault diagnosis system to distinguish different types of faults and provide accurate fault information under complex fault situations. In the proposed system, a bank of hybrid fault observers (HFOs) is designed for the actuator fault detection and sensor fault location, and a nonlinear fault-tolerance estimator (NFTE) is built for the actuator fault estimation. In addition, a switching structure is designed to vary the working process between banks of HFOs and the NFTE. To prove the effectiveness of the proposed approach, a series of simulations was conducted based on a nonlinear aeroengine model. The results verified that the proposed hybrid diagnosis system is capable of locating the sensor fault, detecting the actuator faults, and estimating the actuator faults effectively.

Counterrotating axial fans (CRF) are being adopted in air ventilation to meet the demand for a higher pressure head and swirl-free discharge. This experimental and numerical study is aimed at investigating the influence of blade staggering and interdistance between the front rotor (FR) and the rear rotor (RR) on the aerodynamic performance of a low-speed CRF. The latter was built from a baseline single-rotor fan (SRF) using the same blade profiles. The results reveal the conversion of a large part of the swirl flow into a static pressure, and the total-to-static isentropic efficiency is improved under certain setting parameters of the FR and RR. Moreover, the counterrotation seems to reduce the aerodynamic loading in both rotors blades and to allow a stable operation over a broader flow-rate range. Among the examined configurations of CRF, the optimal has been identified.

Carbon fiber–reinforced polymer (CFRP) composites have been widely used to repair and strength concrete structures. Nevertheless, the durability and long-term performance of FPR-strengthened structures are still not well understood. To this end, nondestructive techniques (NDTs) such as acoustic emission (AE) are usually adopted for the inspection and monitoring of composite structures. The objective of this study is to monitor the damage modes in CFRP-strengthened reinforced concrete structures using the AE technique together with advanced statistical analysis and pattern recognition (PR) methods. Three concrete cube specimens bonded with CFRP sheets and two full-scale RC beams before and after retrofitting were tested to acquire AE data originating from critical damage mechanisms. Because the damage mechanisms in the retrofitted RC beams are unknown a priori, a methodology based on the unsupervised $k$-means clustering analysis, and the supervised neural networks (NNs) were developed. By applying $k$-means clustering analysis, each data cluster was identified to associate with one or more damage mechanisms for the typical specimens. The NN models based on multilayer perceptron (MLP) and support vector machines (SVMs) were then created and applied to other similar samples, which show quite satisfactory performance on damage mode identification.

The involvement of complex manufacturing and fabrication process of composite material has been the main cause of uncertainty. Therefore, the stochastic analysis of flutter characteristics of laminated composite structures is essential. This research work is focused on two different approaches of stochastic analysis of flutter characteristics. First is the perturbation technique, which used the Hessian and gradient of quadratic response surface, optimized based on response surface methodology. The second is polynomial neural network (PNN)-based uncertainty model, which is apparently attempted first in this study. PNN is a group method of data handling–based algorithm developed to relate the input material properties and flutter response. Further, the reliability study is carried out using first-order reliability method (FORM). The outcome of PNN and perturbation technique is compared with Monte Carlo simulation (MCS) and FORM. Perturbation technique is found to be highly efficient and provide results with great accuracy compared with MCS results. Although the PNN-based model showed better computational efficiency compared with MCS, it is not found to be beneficial over perturbation technique. Also, the sensitivity analysis is carried out and the material properties, which greatly affect the flutter characteristics, are identified.