Journal of Aerospace Engineering

This paper investigates two robust finite-time controllers for the attitude control of the spacecraft by using a modified terminal sliding mode surface. With the use of terminal sliding mode control and adaptive control, the controllers under input saturation can compensate both external disturbances and the uncertainty of the model parameters. By using the first and the second controller, the closed-loop system is finite-time stability and practical finite-time stability, respectively. As the system model is based on the rotation matrix, both controllers are without unwinding. Numerical simulations are given to demonstrate the effectiveness of the finite-time controllers.

In recent decades, the exploration of celestial bodies has moved to surface-based landers and rovers designed to perform experiments on the ground. The presented research focuses on legged rovers designed to climb steep slopes covered in uncompacted regolith, similar to those found on lunar crater walls. These rovers would claw into the surface using dynamic anchors on their feet. Physical testing is performed by engaging a series of anchors with lunar simulant. Multiple anchor configurations are tested and evaluated to determine the geometries with the most beneficial characteristics for dynamic anchoring. Several anchor geometries show sufficient holding forces for use on future legged exploration rovers. Additionally, two methods for predicting the anchoring forces with limited knowledge of the regolith are presented, one using discrete elements, the other regression curve-fitting. Both methods show a prediction accuracy for peak anchoring forces well within an order of magnitude.

This paper presents a finite-time adaptive fault-tolerant $H_{\infty }$ control method for a two-dimensional airfoil flutter with parameter uncertainties and external disturbances. Sensor and actuator faults are both considered. Flutter motion equations of an airfoil with cubic hard spring nonlinearity are established based on the trajectory optimization for reentry vehicle. A finite-time adaptive fault-tolerant $H_{\infty }$ controller is designed to deal with the airfoil flutter. The stability of the proposed controller is proved theoretically. Numerical simulations are given to demonstrate the effectiveness of the scheme.

An ultracompact serpentine inlet, which contains a compact diffuser as short as 2.3 times the exit diameter, is experimentally and numerically studied. An internal bump is introduced to redistribute the boundary layer flows and suppress the massive flow separation. It is found that two vortex pairs with the same rotating direction appear at the diffuser exit. The abnormal upper vortex pair helps to avoid massive flow separation at the aft part of the diffuser. According to the calculated flow patterns on the symmetry plane, it is found that a small reversed flow region appears near the lee side of the bump. The skin-friction line plots on the top and bottom surfaces indicate that three-dimensional flow separation occurs. Three vortex tubes are obtained in the diffuser by the Q-criterion. The first, which develops along the spanwise direction downstream of the bump, induces the reversed flow. The second and third vortex tubes, which originate from the sides of the top and bottom surfaces of the duct, collect the low-momentum fluids and transport them downstream spirally. At a test Mach number of 0.70, the total pressure recovery is 0.987 and 0.944 at exit Mach numbers of 0.253 and 0.491, respectively.

The mixing efficiency of a shifted limiting tab of a rectangular cross section, with corrugations all along its edges, placed downstream of a circular Mach 2 nozzle operated at different levels of expansion corresponding to nozzle pressure ratios (NPRs) 3–8 in steps of 1, has been studied experimentally. The corrugation geometries studied in this investigation are circular-arc and triangular. The corrugated tabs offered a geometrical blockage of 5% of the nozzle exit area for all the cases studied. The shifted positions studied are 0.25D, 0.5D, and 0.75D. In addition to shifted tabs, tabs at the nozzle exit (0D) are also studied for comparison. For the Mach 2 jet, the nozzle pressure ratios of 3–8 correspond to a range of pressure ratios $p_e/p_a$ from 0.383 to 1.022. Thus, these NPRs establish adverse and favorable pressure gradients at the nozzle exit. The results of this study show that the mixing promotion caused by tab at 0.75D is the best, resulting in a core length reduction of approximately 86% in the presence of a marginally favorable pressure gradient of 2.2%. Comparison of the present results with the corresponding results for uncorrugated limiting tabs available in literature show that the mixing promotion caused by the corrugated tab is superior to the tab without corrugation.

This paper presents the design and analysis of a robust neural adaptive backstepping control (RNABC) for an autonomous quadrotor helicopter perturbed by time-varying external disturbances. The work uses a backstepping method and a radial basis function neural network (RBFNN), which estimates perturbation. A gravitational search algorithm (GSA) is included to optimize the backstepping controller for a nominal dynamic model of a helicopter. Disturbances caused by external sources are estimated using the global estimation attribute of the RBFNN. To further improve the control design performance, a robust compensator is introduced to eliminate the approximation error produced by the neural approximator. Asymptotical stability of the closed loop control system is analytically proven via the Lyapunov theorem. The main advantage of the proposed methodology is that it requires no advance knowledge of the disturbances. A quadrotor helicopter is simulated to track trajectories. The effectiveness of the controller is also validated by realistic effects such as model uncertainty and measurement noise. The results show the efficiency and usefulness of the designed approach.

The paper studies a method which uses a solid propellant microthruster array (SPMA) as an actuator to control vibration of solar panels and the attitude of a satellite simultaneously. To solve the controller design problems of the system with periodic switching characteristic, which is caused by the mismatch between the solid propellant microthruster combustion duration and the digital control system sampling period, Floquet theory is used to obtain an equivalent time-invariant model. Based on the state-dependent Riccati equation method, an optimal antisaturation controller is designed to overcome the saturation problem of the attitude flywheel which occurs in the cooperation with the SPMA actuator. In terms of the input quantization problem of the SPMA actuator, a switching strategy is proposed to achieve global asymptotic stability. Comparison of the numerical results for SPMA and piezoelectric actuators show that the SPMA actuator can suppress the low-frequency vibration much more quickly. Finally, a Monte Carlo approach is used to investigate the robustness of the controller and the switching strategy with respect to model uncertainties.

The paper presents the aerodynamic optimization of a morphing leading edge airfoil using a parameterization based on the class/shape transformation (CST) technique associated with a dedicated procedure to keep the arc length of the curve constant in order to limit the axial stress of the deformed shapes. The optimization is performed with a standard methodology based on genetic algorithms, comparing the results for three different aerodynamic models. Whereas the solutions obtained with the third model are standard droop nose shapes, those found via transitional models show an uncommon deformation with an upward leading edge deflection. A metamodel-assisted optimization loop is used to solve a known problem, showing that an artificial neural network is able to provide a reduction of the convergence effort when approximating the highly nonlinear relationship between the constant arc length parameterization and the aerodynamic behavior predicted with two of the models.

Stratospheric airships can perform long-duration, wide-area observation missions above a specific location. The conceptual design of streamlined stratospheric airships is researched in this paper. Multidisciplinary analysis models for stratospheric airships are established, including geometry, aerodynamics, propulsion, and renewable power systems. A conceptual design method focusing on the energy balance for stratospheric airships is proposed, taking minimum total mass as the objective function. A conceptual design of a stratospheric airship for area observation was achieved, and the feasibility of the design results was verified. In addition, the influence of the wind resistance strategy on the conceptual design was analyzed, and simulation results showed that, by adjusting power consumption during day and night with a maneuverable wind resistance strategy, the total mass and renewable power system mass of the stratospheric airship can be greatly reduced.

A magnetic sampler for future sample return missions on asteroids was developed. The sampler had a simple configuration, consisting of a solenoid coil, direct current (DC) power supply, and switching circuit, and the sampling was conducted using the coil gun mechanism. The sampler did not require mechanical moving parts, gas, or liquid, and its control was simple; thus, it was reliable for use in a space environment. The authors analyzed the particle motion in a magnetic field on Earth and in space environments; the sampler performance was evaluated by numerical calculations on the basis of the distinct element method and experiments in air and vacuum environments. The results of numerical calculations and experiments were consistent qualitatively with the analytical calculations. The magneto-motive force and its applied time mutually affected the dynamics of particles. The magnetic interaction of particles in the magnetic field also affected the sampler performance. Moreover, the sampler performance in vacuum and a microgravity of 0.00001g was better than that on Earth owing to the absence of air drag and small gravitational force.

The Kevlar 49 composite material (DuPont, Wilmington, Delaware), which is of light weight and high strength, is used increasingly as a containment structure for fan blades. This paper proposes a spinning impact testing method to investigate the containment ability of Kevlar 49 composite case after fan-blade-out (FBO) events. Three spinning impact tests are conducted to closely simulate actual FBO events by considering the impact angle and blade rotating factors. These tests show three distinct results: contained, critical contained, and uncontained. The computational model is developed and correlated with the experiments. It shows that eight layers of Kevlar fabric can succeed in catching a released blade with 4,000 J energy under spinning status, of which 3,083 J is absorbed by the Kevlar fabric. A comparison with the data published in the Federal Aviation Administration reports indicates that Kevlar 49 composites have greater containment potential under a spinning impact. A different containment mechanism is found in which a bulge is formed in the radial direction of the fabric, flowing along the circumference together with the rotating blade when the contact zone is expanded. This mechanism should be considered when applying woven fabric to an aero-engine containment system.

The vibration characteristics of composite pyramidal truss core sandwich plates with piezoelectric actuator/sensor pairs were investigated, and the active vibration control methods of the structural system were also developed. The face sheets and truss core were all made of carbon fiber–reinforced composites. In order to construct an effective control system, the piezoelectric materials were symmetrically bonded on the top and bottom surfaces of the sandwich plate to act as the actuators and sensors. Hamilton’s principle with the assumed mode method was used to establish the equation of motion of the composite pyramidal lattice sandwich plate bonded with piezoelectric materials. The natural frequencies of the composite sandwich panel were calculated and validated by the finite-element method. A negative velocity feedback control method and a linear quadratic regulator (LQR) were employed in the controller design. The controlled vibration responses of the composite sandwich panel with the two different controllers under transverse impulse excitation were calculated. Numerical results show that the proposed active control methods can effectively suppress the vibration of the composite pyramidal truss core sandwich panel, and that LQR control requires less energy than velocity feedback control.

This paper presents a novel dynamic scaling–based non-certainty-equivalent adaptive control method for the attitude tracking control problem of spacecraft in the presence of inertia uncertainties. By virtue of the modified dynamic scaling factor and an angular velocity observer, the proposed controller provides a more direct approach to circumvent the integrability obstacle of the basic immersion and invariance (I&I) adaptive method without any information about the bounds of system uncertainties. Compared with the existing filter-based non-certainty-equivalent adaptive attitude controller, the proposed controller achieves significant order reduction in closed-loop systems and removes the constraint on initial values of the parameter lock property. The dynamic gains in the proposed method are modified to have linear growth in the dynamic scaling factor and restricted by the given upper bound in order to avoid bad transient performance. Numerical simulations are presented to highlight the superiority of the proposed method compared with the filter-based I&I control method.

Deorbit guidance laws steer a reentry spacecraft from the parking orbit to the entry interface (EI) while satisfying some terminal constraints. The guidance precision has great influence on the shape of the reentry trajectory, heat flux, aerodynamic overload, and safety. When the Earth’s $j_2$ effect is taken into consideration, the guidance law becomes either complicated or of heavy computational load. Therefore this paper proposes a velocity-to-be-gained deorbit guidance law as a simple and precise solution. The main portion of the velocity to be gained is obtained from solving a two-body two-point boundary value problem, and the correcting portion related to the $j_2$ term is calculated analytically through the state space perturbation method (SSPM). The SSPM uses a linearized perturbation motion equation derived in local vertical, local horizontal coordinates. A numerical simulation is conducted to verify the validity and accuracy of the proposed method. The guidance law can satisfy terminal constraints on flight path angle and location of the EI and can be generalized with little effort to cases in which other terminal constraints are specified or Earth’s higher-order nonspherical gravity is considered.

This paper focuses on the seismic control design of a new type of shape memory alloy (SMA) damper named double X-typed SMA (DX-SMA) damper. It is installed in a frame-typed bridge pier. A parametric study based on time-history analyses is carried out to investigate the influence of the strength ratio and the span ratio. The amplification effect of the DX-SMA damper is tested by a low cyclic loading process. The performance of the structural safety and the self-centering device postearthquake are evaluated by analyzing the residual top displacement and the maximum top displacement of the frame pier. Lastly, the energy dissipation of the SMA damper is investigated. Design recommendations are suggested following the results of the parametric study.

The Ultrasonic/Sonic Driller/Corer (USDC), a novel percussive drilling mechanism, is a tool for rock sampling in support of planetary exploration. This mechanism consists of three components: an ultrasonic actuator, a free mass, and a drill bit. The free mass, bouncing between the horn tip and drill bit end, delivers energy from the transducer to the surface of rock. This paper presents an analysis of the collision between the free mass and horn tip/drill bit end. The collision between different components is numerically analyzed by an elastic-dynamic approach, and finite element analysis is used to verify the calculated results. A simplified integrated model is programmed to simulate the operation of this device. Distinct features of this device, i.e., the failure and optimum states, are obtained from the integrated model and experiments. It is observed that single and multiple collisions may occur between the free mass and horn tip in one collision cycle. A decrease in the vibration amplitude makes multiple collisions and acceleration of the free mass impossible. Moreover, the failure and optimum states are affected by the weight of free mass and preload.

Mechanical properties of the polymer composites are rate-dependent. The representative volume element (RVE) of unidirectional composites considering the rate-dependent effect of polymers is established with the Improved Bodner-Partom model implemented through three different numerical integration methods. Then, to increase stable time increment, the size of the RVE model is increased properly, and to relieve stress oscillation in the model, a mass scale method is adopted. Through applying periodic boundary conditions, one element model and one RVE model with fiber square arrangement are used to verify the accuracy of a constitutive model of the polymer and simulation method of RVE, respectively, with simulation results compared with the experiment results. The corresponding CPU time consumed by the three different integration methods is compared simultaneously. The models with fiber random or regular distribution (square, diamond, and hexagon arrangement) are analyzed to reveal the effects of fiber arrangement on predicted response under high strain rate loading. Effects of void volume fraction on the response predicted from fiber regular arrangement models are presented. A numerical scheme is developed to establish models with fibers and voids random distribution. With Python language, effects of the fiber arrangement on the predicted results and influences of existence of voids with random distribution on composite mechanical properties are revealed using the Monte-Carlo method.

The protection cover is an important equipment installed in the main landing gear cabin to protect the hydraulic lines and equipment. To assess the safety of the protection cover, the authors conducted an experimental and numerical study on the dynamic responses of protection cover in the main landing gear cabin under blast pressure of aircraft tire blowout, and the influences of various blast conditions were discussed. In the experimental study, three protection covers with different burst target points and tire inflation pressure were tested. Based on the data recorded by the pressure transducers, a new blast function in the time and space domains was proposed to describe the pressure distribution on the outer surface of the protection cover during the process of aircraft tire blowout. To validate the proposed burst model and finite-element model of protection cover, the simulations results were compared with the experimental results. It shows that the numerical calculations and experiment results exhibit a satisfactory agreement, which indicates that the models in the present study are reliable and can be used for further studies. Moreover, considering the influences of the burst target point and the tire inflation pressure, the dynamic responses of the protection cover were analyzed in detail with the finite-element method. According to the numerical simulation results, when at a certain inflation pressure, the protection cover is found to undergo the severest impact with the burst target point A, which is close to the center of the protection cover.

A wheel-terrain interaction mechanics model based on terramechanics and simplified force and moment reconstruction is introduced. The model comprises pressure-bearing parameters (combined deformation modulus and slip sinkage exponent), wheel load, driving torque, slip ratio, and wheel sinkage. A genetic algorithm is used to identify the pressure-bearing parameters using a calibration dataset. Experiments were performed on the calibration dataset and a prediction dataset using a single-wheel soil bin to measure the driving torque, wheel sinkage, and drawbar pull of a griddle net wheel under different slip ratios. The proposed model comprises two processes. The first process is to determine the slip sinkage exponent using its empirical value to estimate the combined deformation modulus, and the second is to correct the slip sinkage exponent in which the value of the combined deformation modulus is fixed. The prediction dataset was used to validate the model, with the wheel sinkage and drawbar pull calculated using the wheel-terrain interaction mechanics model with the identified pressure-bearing parameters. The correlation coefficients of the predicted and measured values of the wheel sinkage and drawbar pull were 0.9712 and 0.8902, respectively. An additional experiment was performed with a continuous slip ratio ranging from 0.16 to 0.7 and a wheel load of 50 N to validate the proposed model, and the correlation coefficient between the predicted and measured wheel sinkage was 0.9660. The experimental results demonstrate that the genetic algorithm is able to accurately predict the pressure-bearing parameters, which can be used for traversability prediction, risk assessment, and automatic path planning.

A reentry guidance method for low lift-to-drag (L/D) ratio vehicles in a lunar return mission is presented. Different from the reference drag tracking scheme, method in this paper tracks a weighted profile of altitude and downrange using a command filtered backstepping approach. First, the dynamics of the tracking error are established with dispersions and uncertainties explicitly taken into consideration. Then, in accordance with the backstepping technique, the weighted reference profile is tracked with the help of a virtual control signal. Next, by tracking the virtual control signal, a real control input is derived. Dispersions and uncertainties are accurately estimated by a well-designed disturbance observer and compensated for by the estimates. The derivatives of the virtual control needed by the backstepping algorithm are obtained by filtering the virtual control signal with a second-order command filter. Finally, the stability of the whole system is analyzed, and extensive numerical simulations are conducted to demonstrate the effectiveness and robustness of the proposed method.

This paper describes a fault-tolerant decoupling-control algorithm for spacecraft incorporating single-gimbal control moment gyroscopes (SGCMGs), simultaneously considering the SGCMG rotor and gimbal faults. Double-loop control theory is utilized to design the attitude system. An outer-loop controller is designed to obtain a control torque with proportional-derivative-type (PD-type) technology, and a singular direction-avoidance (SDA) steering law is adopted to calculate a virtual gimbal-rate vector, which will be the reference signal for the inner-loop controller. The actuator fault is not considered in the outer-loop system. In the inner-loop system, an active-disturbance rejection controller (ADRC) is designed. The ADRC incorporates an extended-state observer (ESO) to estimate the total disturbance for each SGCMG gimbal or rotor in order to track the virtual gimbal rate or nominal angular momentum by considering the gimbal- and rotor-rate faults. In this way, a decoupling controller with fault-tolerance capability is achieved. The simulation results demonstrate that the proposed method is strongly robust against actuator faults.

A dynamic model is proposed and validated to analyze the landing performance of a vertical landing reusable launch vehicle (RLV) with a novel oleo-honeycomb two-stage damper. Based on this model and design requirements, seven critical landing conditions are found as a foundation of the following multiobjective and multiconstraint optimization for oleo-honeycomb damper. In order to overcome the time-consuming and variable coupling problem of the optimization, a combinatorial optimization method is proposed and verified with collaborative optimization (CO) combined with an archive-based micro-genetic algorithm (AMGA). Meanwhile, relaxation factors and the radial basis function (RBF) model are employed to improve the computational accuracy and efficiency. After obtaining the optimal solution, the influences of the landing strut friction coefficient and ground friction coefficient on the soft-landing process are analyzed. The results show that the combinatorial optimization method is feasible and efficient, and the selection of honeycomb crushing force and hydraulic damper’s parameters affect and restrict each other. The oleo-honeycomb damper has some advantages over traditional dampers and two factors of initial landing conditions have different influences on touchdown stability, impact acceleration, and nozzle ground clearance.

The response of soil–structure interfaces is of primary importance for the construction and design of infrastructure because all loads, whether static or dynamic, are transferred to the ground through these interfaces. This paper presents a laboratory investigation that characterizes the interface shear response of two lunar simulants, JSC-1A and GRC-3, and one Martian simulant, JSC-Mars1. Through a series of axial and torsional interface shear tests, this investigation thoroughly studies the effect of continuum surface roughness magnitude on the interface response and demonstrates that the JSC-1A simulant mobilizes the largest interface strength, followed by JSC-Mars1 and then GRC-3. Furthermore, the results indicate that torsional shearing against regolith mobilizes larger peak interface friction angles than those mobilized during axial shearing. A methodology for isolating the interface friction response from the passive resistance components is presented, and it is shown that the passive resistances are responsible for the difference in mobilized axial and torsional interface strength. A discussion of the implications of interface response on space exploration, as well as extraterrestrial immobility and infrastructure development, is included.

In this work, to demonstrate the efficiency of model reduction in design optimization, a parametric reduced-order model (PROM) was adopted in conjunction with an aeroelastic optimization process. Flutter speed was chosen as an objective function, and structural properties (material density, Young’s modulus, and Poisson’s ratio) as well as fluid properties (air density) were defined as the design variables. The flutter calculation was performed for a Goland wing, using finite-element modeling for the structure and the vortex lattice method for the aerodynamic part. A gradient-based optimization technique and a global optimization method were used to seek a maximum flutter speed. Comparison of optimization results from the full-order model (FOM) and PROM shows that the proposed optimization process yields the same optimum flutter speed as the FOM and yet reduces the computation time significantly, by up to a factor of four.

To achieve tracking performance and robustness simultaneously during an aircraft’s autonomous short landing phase, a three-dimensional automatic landing system designed using robust adaptive backstepping control is investigated to direct aircraft glide-slope and flare maneuvers. To avoid the difficulty of analytically calculating the virtual command derivatives, a series of virtual command filters are introduced. By confining the input of the command filter, the flight states are subject to the landing constraints. To guarantee tracking performance in the presence of input and state constraints, antiwindup compensators are constructed by the saturation errors. To approximate the compound uncertainty brought by the ground effect and aerodynamic uncertainties, nonlinear disturbance observers are employed to estimate the uncertainties within a finite time. Synthesized with backstepping and adaptive proportional-integral-derivative control, the automatic landing control law is designed. Stability analysis shows that the closed-loop system achieves ultimately asymptotically bounded stability under weak uncertainties. Numerical simulations of autonomous landing demonstrate that the proposed control law can achieve both landing trajectory capturing performance and robustness against uncertain aerodynamics.

This paper describes the study of unsteady forces acting on a two-dimensional morphing airfoil under the assumption of incompressible, inviscid flow. Through conformal mapping, the flow over a morphing airfoil was mapped to the flow over a movable, variable radius circle, and a mathematic analytic model of unsteady lift was derived. To verify the unsteady lift model, it was then simplified for the case of a thin airfoil whose camber changes with time, and numerical simulations were conducted for the flow over the deformable thin airfoil. The unsteady lift caused by airfoil deformation is equivalent to the quasisteady lift added to the corresponding additional unsteady lift. The additional unsteady lift is linearly correlated with the first time derivative and second time derivative of the airfoil surface displacement. When the airfoil surface undergoes a vibration with a large reduced frequency, the additional unsteady lift is mainly decided by the second time derivative of the surface displacement.

This study applies structural dynamics monitoring to identify the presence of bridge scouring through tracing the changes of vibration characteristics induced by scour. Four scour-indicating dynamic features, i.e., the frequency change ratio, modal assurance criterion, mode shape curvature, and flexibility-based deflection, are applied as the scour indicators. Based upon the study of a cable-stayed bridge, each scour indicator is calculated and its corresponding relationship with scour depth is obtained. The feasibilities and effects of using each indicator ae comprehensively examined. After comparison, the flexibility-based deflection is finally recommended as a sensitive and practical indicator to identify the presence of the scouring. When applied in practice, the deflection can first be extracted from the ambient vibration measurement; then the scour depth can be back-deduced based on the preobtained relationship between the deflection and scour depth. For the specific cable-stayed bridge investigated in this study, the pylon-based deflection indicator can identify the scour presence more effectively and distinguishably. The scour location can also be known because the scoured pylon or pier shows many obvious deflection changes compared with those without scour effects. This study demonstrates that tracing the change of vibration responses can both qualitatively and quantitatively identify the scour of cable-stayed bridges. This vibration-based technique does not require any underwater devices or operations, which could be integrated into a routine assessment task for bridges.

This paper is aimed at investigating the risks associated with flying qualities of a tilt-rotor aircraft slung-load system. Taking the XV-15 tilt-rotor aircraft (Bell Helicopter Textron, Fort Worth, Texas) as an example, a coupled nonlinear mathematical model is first introduced red and trimmed, where a mass point model is used for slung load to study its effect on a tilt-rotor aircraft slung-load system. Second, the linearization of the previous nonlinear equations by small perturbation hypothesis is performed for open-loop frequency-domain response analysis. Finally, based on the flying qualities requirements for military aircraft, the results of tilt-rotor aircraft with and without slung load are thoroughly compared. The results presented are of significance for engineering reference because there are few open publications available to date on tilt-rotor aircraft slung-load system trims, controllability, and flying qualities. Results indicated that due to the slung load, the trim results of collective pitch increase significantly, the lateral response in hover is significantly affected in the lower frequency range, and in helicopter mode some flying qualities get worse while the others are slightly changed.

Ice accretion can considerably alter aerodynamic characteristics of bridge cables. The present study focuses on flow characteristics in the cable wake for three arrangements of ice accretion on the cable surface. This is studied for the subcritical flow regime (subcritical Reynolds numbers) at various downwind distances of the cable. The analysis is based on velocity and force measurements carried out in a boundary-layer wind tunnel. Two different bridge-cable models are used, i.e., smooth cylinder representing dry cable, and artificially roughened cable representing ice-accreted bridge cable. The effects of ice accretion on flow characteristics in the cable wake are particularly exhibited when ice is oriented laterally downward for the cable section mounted horizontally in the wind tunnel. This is characterized by strong velocity reduction and turbulence enhancement in the cable wake.