The Aerodynamics of Heavy Vehicles: Trucks, Buses, and Trains

Recognizing both the pollution effects of fuel use and the likely increases of fuel cost in coming decades puts high priority on alternative energy for trucks, buses and trains. There are still gains available in decreasing aerodynamic drag and rolling friction, using efficient engines, and minimizing fuel waste, but it is appropriate to explore decisions that would be suitable if fossil fuel were deemed unattractive. One then would consider utilizing natural energy (sunlight, wind, wave), getting energy from braking, employing hydrogen, putting different priorities on trucks vs. buses vs. trains, exploring integration with water deliveries and automatic local air transport, etc. Such an investigation might illuminate early alternatives that would at least permit partial improvements.

The aerodynamics development of commercial vehicles has evolved over many years. Sixty-five years ago, the Labatt Brewing Company developed a streamlined truck for advertising purposes and to provide larger capacity and higher cruising speeds, Figure 1. The success of this effort is demonstrated by the fact that while trucks of the day travelled at 35 mi/h, the Labatt truck could cruise at 50 mi/h with a fifty percent larger load. The focus today is no longer on speed, but on energy conservation. It is beneficial for a country to minimise its energy utilisation and equally beneficial for its trucking industry to make money while doing so.

Proposed in 1997, DES was applied to an airfoil beyond stall in 1999, and then to a range of bluff bodies. Its accuracy has often been far superior to that of steady or unsteady Reynolds-averaged Navier-Stokes methods, and it avoids the Reynolds-number limitations that plague Large-Eddy Simulation. Cases fall into three classes: simple shapes such as cylinders and spheres; transportation components such as landing gear, simplified; and full airplane geometries. All are manageable on present computers, some even on personal computers. Simple shapes now and then yield surprises, but DES appears sound and reacts well to the type of boundary-layer separation (i.e., laminar or turbulent), and to grid refinement. However, it is possible to confuse the method by using a grid density that is both too fine for RANS and too coarse for LES. Component studies display progress, without reaching an industrial level of accuracy in predicting all forces. The few full-airplane predictions have been successful, thanks to high CPU power, and partly thanks to fixed separation lines.

The results of the large eddy simulation of the flow around a simplified bus presented in [11, 12] are used to describe this flow in detail. Using timeaveraged trace lines on the surface of the body, the patterns of the shear-stress lines are revealed and used to identify bifurcation lines and critical points (zero-shear-stress points) in the flow. This information is then used to establish a complete picture of the flow on the surface of the body that can be used for understanding soiling and accumulation of water on the surface or in determinations of aeroacoustic noise sources. Kinematical investigations of the flow in two symmetry planes were done to reveal the critical points in the flow. With this it was proven that the flow resulting from numerical simulation is kinematically possible.

Reynolds-Averaged Navier-Stokes simulations of the flow around circular cylinder with V-shaped longitudinal cavities are carried out to study the effect of the cavity geometry on the flow characteristics. In particular the effect of the cavity depth on the unsteady aerodynamic forces is analyzed. It is found that the cavities reduce the overall drag and the amplitude of the lift fluctuations.

Complex CFD for everyday analysis encompasses quick, efficient grid generation and robust, accurate flow solution. These two disciplines — while distinct and possessing disparate constraints — contribute to produce a solution that is limited by the proper application of the selected tools. With simple forethought, the best assets of each discipline can be combined to further enhance the impact of the other. The fusion of unstructured grids and Detached-Eddy Simulation (DES) enhances the optimal benefits of each to provide an accurate, efficient, and robust computational model. The ability of unstructured grid technology to model complex geometries quickly and efficiently is further highlighted by the prism boundary layer and nearly isotropic cells that it creates outside the boundary layer. The tightly clustered boundary layer lends itself to traditional RANS techniques while the isotropic cells clustered in regions of separated flow improve and elevate DES to its optimum performance. Grid adaption has shown promise in a similar manner. By reducing the cell count in benign regions and increasing cells in turbulent regions, convergence is accelerated and accuracy is improved. Time-accurate DES flow solutions obtained with Cobalt will be presented and compared with experimental data. Vehicle analyses — both aircraft and automotive — will be examined and compared to traditional numerical methods for turbulence.

The automotive industry has a high demand for reliable simulation methods capable of tackling the complex turbulent air flow around vehicles. The Ahmed reference model is a generic car-type bluff body with a slant back. It is frequently used as a benchmark test case for this kind of flow. In spite of the relatively simple geometry of the Ahmed body, the flow around it retains some main features of the flow around real cars.

Predictions of the flow around the Ground Transportation System (GTS) are obtained from Detached-Eddy Simulation (DES) and from the Reynolds-averaged Navier-Stokes (RANS) equations. Two methods are compared with experiment, but only one grid is used, as the work is in progress. Grid variations will be needed. The computations are performed at a Reynolds number based on body width and inlet freestream velocity of 2 × 106, at 0° and 10° yaw. Solutions are obtained using unstructured grids comprised of approximately 6 × 106 elements (prisms and tetrahedra) and the commercial flow solver Cobalt. No attempt is made to model laminar-to-turbulent transition. Instead, the predictions are of the fully turbulent solution obtained via prescription of a small level of eddy viscosity at the inlet to the computational domain. The RANS predictions are of the steady-state solution using the Spalart-Allmaras model. For 0° yaw the predicted drag coefficient Cd from the RANS of 0.370 is substantially larger than the measured value of 0.249. DES yields a more complex and three-dimensional structure in the separated regions. One of the improvements over RANS is a more accurate prediction of the back pressures, resulting in the DES prediction of Cd = 0.279. At 10° yaw, DES predictions of the body-axis drag are again closer to measurements than obtained using RANS, but substantially larger than the measured value. One source of the discrepancy is a more significant separated region near the front leeward corner than observed in experiments, resulting in tangible differences in the pressure distribution along the lee side.

The paper assesses the use of the most popular standard k-ε model and the most accurate Reynolds-stress model for various simple and complex flows including vehicles, and their impact on steady and transient RANS calculations. At the same time, the error of the steady state approach for the flows where transient effects are important, is analyzed. The Hybrid Turbulence Model is examined as an alternative solution, e.g. the recent proposal of combining the Boussinesq’s concept with the second moment closure. Finally, new possibilities in the further use of steady and averaged transient results will be addressed e.g. acoustic calculations. The paper compiles previous and present work performed at AVL List GmbH using the in-house commercial CFD software AVL Swift.

The study of the aerodynamics of ground-vehicles has greatly benefitted from computational fluid dynamics (CFD). However, in practice, there are some hurdles that remain to be overcome before CFD can be fully established as a design tool. The paper will review some of the recent developments in the CFD modeling of the subject flow under several important themes outlined below.

The presentation includes the use of the computational Fluid Dynamics (CFD) at Kenworth Truck Company in the last few years, how do we identify projects to build confidence in CFD simulation, follows by the discussion on the aerodynamic validation process such as wind tunnel testing, proving ground on-track testing and over-the-road real world testing. A brief look at the history of the concept and development process and how we integrate CFD in the current development design process, identify areas of CFD development to improve the simulation and testing results correlation in the future.

Direct numerical simulations (DNS) via Navier-Stokes equation is adequate for Newtonian fluid flows at macroscopic scales including high Reynolds number turbulence. On the other hand, we argue that the large eddy simulations (LES) of turbulent flows is better achieved physically via Boltzmann equation based kinetic formulations and the lattice Boltzmann method (LB). Among other features compared to the Navier-Stokes based approach, the major differences in the alternative approach include 1) Realization of higher order turbulent eddy effects; and 2) Realization of physical boundary conditions. As a numerical method, LB has shown to be advantageous in doing LES involving complex geometries and flow fields, and it allows for accurate and efficient simulations of time dependent turbulent flows with potentially relatively simpler turbulence models. In this presentation, the fundamental concept of the kinetic based approach for LES will be introduced. Various numerical results on basic turbulent benchmark flows as well as on practical engineering flow problems will also be shown.

The effectiveness of CFD for vehicle aerodynamic design depends on a number of factors, notably: accuracy, cost, turnaround time and ease of use. Commercial CFD software developers have traditionally rightly placed strong emphasis on the first three, with considerable success, as will be illustrated by examples in this presentation. These achievements have led to increased use of CFD in the aerodynamic design process, but the levels of skill and experience required are arguably now becoming the limiting factor on its exploitation. A way around this problem, using the Expert Systems approach, is presented.

The turbulent flow structure in the near wake of a pickup truck model has been investigated experimentally using Proper Orthogonal Decomposition (POD) analysis of Particle Image Velocimetry (PIV) data. The experiments were conducted in the 2’x2’ wind tunnel at the University of Michigan at Reynolds numbers based on model width of 3x105. A model of a pickup truck with extended cab, 432 mm long by 156 mm wide by 149 mm tall, was used. PIV measurements of the velocity field in several planes of the wake including the symmetry plane were obtained using a large sample size, sufficient to determine the mean flow and the Reynolds stresses in the wake. The mean flow structure in the symmetry plane consists of separated shear layers originating at the edge of the cab and from the underbody flow. For this particular geometry, there is a recirculating flow region behind the cab ending upstream of the tailgate; but there is no mean recirculating flow region behind the tailgate. The mean flow pattern in a horizontal plane behind the tailgate suggests that this is due to trailing streamwise vorticity. POD analysis of the data provides useful information on the unsteady large scale structures in the wake. It is shown that only a few orthogonal modes (~20) contain a large fraction of the fluctuation energy (~60%), as expected. But, more important, the analysis isolates features of the unsteady large scale turbulent structures into different uncorrelated modes. It is shown that the development of vortex shedding in the underbody flow shear layer is described by a few modes, while oscillations of the recirculating region are captured by other modes. This feature of the analysis makes it very useful to the study of the structure and dynamics of complex bluff body wakes.

The quantification of experimental flows is a problem that poses several challenges, the most obvious of which is how to extract motion from an “invisible” phenomenon. In general, flows can be analyzed through a sequence of still images (Singh 1991). For example, the motion of patterns generated by dye, clouds or particles can be used to obtain such a time sequence of still images. The main problem with using a continuous-intensity pattern, generated by scalar fields (e.g., dye patterns), is that it must be somehow discretized and contain variations of intensity at all scales before mean and turbulent velocity information can be obtained (Pearlstein and Carpenter 1995). In this respect, the discrete nature of images generated by seeding particles has made particle tracking the method of choice for whole field velocimetry. Displacement and, thus, velocity information can be extracted through statistical methods and other methods such as particle tracking. The spatial resolution of this method depends on the number density of the particles.

This work provides an overview of the technique of Molecular Tagging Velocimetry (MTV) and some of its automotive applications. The various elements of MTV implementation are briefly described in terms of the available molecular tracers, methods of tagging, detection, and processing schemes. The automotive applications of this velocimetry technique are demonstrated in mapping the velocity field of the intake flow into a “steady flow rig” model of an internal combustion engine and flow mapping of cycle-to-cycle variation in late compression of a motored IC engine.

Despite the increased use of numerical simulations in the development and performance optimization of aerodynamic vehicles wind tunnel tests still are of fundamental importance in the related engineering design process. In order to keep the cost down and to increase the data return from the expensive measurement campaigns, there is an interest in improving and expanding the diagnostic tools available.

The 1/8-scale Generic Conventional Model (GCM) was studied experimentally in the NASA Ames 7- b 10-Ft Wind Tunnel. The model was designed for validation of computational fluid dynamics (CFD) for a realistic model of a conventional tractor-trailer vehicle with the engine in front of the cab. The model was simplified so that mirrors, exhaust stacks, flow through engine compartment, and other details were not defined on the model. The gap between the tractor and trailer could be varied from 40 to 80 inches full scale. The model was tested with and without wheels. Side and roof extenders were tested that enclosed the gap from 30 to 60 percent. Two trailer configurations (conventional and lowboy) were tested. Aerodynamic boattail plates were also tested on the back of the trailer for both configurations. A simplified configuration was also tested were the gap between the tractor and trailer was filled in. Experimental measurements included the body-axis forces and moments on the total vehicle, the body-axis drag and yawing moment of the tractor, surface pressures on the vehicle, unsteady pressures on the rear of the tractor and the front and rear of the trailer, oil-film interferometry to measure skin friction, and 3-D particle image velocimetry (PIV) in the gap and behind the trailer. For the basic model, a strong hysteresis effect was observed on the aerodynamic forces and moments with the drag changing over 35% in the loop at yaw angles greater than 10°. The wind-averaged drag coefficients for the model without and with side extenders were 0.594 and 0.437, respectively. Besides reducing the drag, the side extenders also eliminated the aerodynamic hysteresis. PIV measurements in the gap between the tractor and trailer indicate the strength of the vortical flow in the gap was significantly weaker with the side extenders installed. In the lowboy trailer configuration with side extenders, the wind-averaged drag coefficient was 0.376. For the conventional and lowboy trailer configurations with side extenders, the wind-averaged drag coefficients were 0.397 and 0.309, respectively. The methodology for calculation of the force and pressure coefficients is included to facilitate comparison between computation and experiment.

Particle Image Velocimetry (PIV) measurements were acquired in the wake of the trailer and in the gap between the tractor and trailer of the Generic Conventional Model (GCM) truck for the US Department of Energy. The data will be used both for validation of computational fluid dynamics (CFD) codes and for understanding the flow physics. The GCM is a 1/8th-scale, moderate-fidelity model of a full-scale truck. The test was performed in the Army/NASA 7 × 10 wind tunnel at NASA Ames Research Center. Surface pressure and force measurements were made prior to the PIV measurements. PIV measurements were made at two yaw angles and at three horizontal planes for three model configurations, each at a free-stream velocity of 52 m/s (Mach 0.15), which corresponds to a Reynolds number of 1 × 106, based on the width of the tractor. This paper discusses the PIV system, samples of flow data and some of the observed features that may have contributed to the measured drag.

Wind tunnel experiments on the aerodynamics of tractor-trailer models show that the drag on the model is sensitive to the width of the tractor-trailer gap (G) and to the angle of yaw with respect to wind direction. At zero-yaw, relatively low drag is measured up to a critical gap width G/√A ≈ 0.5, where A is the cross-sectional area. At the critical width the drag experiences a sharp and large increase; most of the drag contribution is attributed to the trailer alone. As the gap is widened further, tractor and trailer become increasingly decoupled from each other and the drag reaches a near-plateau, rising much more gradually.DPIV measurements in horizontal planes in the gap show that the flow is steady and consists of a relatively stable, symmetric toroidal vortex when the width is below critical. The symmetry breaks down at the critical gap, as evidenced by intermittent ejections of flow from the cavity to either side of the model. These ejections are believed to be at the origin of the sharp increase in trailer drag. As the gap width is increased further, the nature of the flow transitions from cavity-like to wake-like.These observations can be qualitatively extended to moderate yaw angles (up to ~4 degrees), but the size of the critical gap width diminishes with yaw angle. At higher angles, the drag rises much faster with gap width.The second part of this paper discusses the drag savings that can be realized by arranging two truck-like models in a tandem. Four tandems were formed by combining two models; each of the models was either “rounded” (i.e. lower drag) or “blunt” (higher drag). The drag of any tandem is generally lower than the sum of the drags of the models in isolation. However, the drag savings also depends on the choice of models (rounded vs. blunt) and on which model is placed in front. A rounded model followed by a blunt model achieves the most relative drag savings, while reversing the order produces the tandem with the least savings.

Steady-state Reynolds-Averaged Navier-Stokes (RANS) simulations are presented for the three-dimensional flow over a simplified tractor-trailer geometry at zero degrees yaw angle. The simulations are conducted using the SACCARA multi-block, structured CFD code. Two turbulence closure models are employed: the one-equation Spalart-Allmaras model and the two-equation k-ω model of Menter. The discretization error is estimated by employing two grid levels: a fine mesh of approximately 20 million grid points and a coarse mesh of approximately 2.5 million grid points. Simulation results are compared to the experimental data obtained at the NASA-Ames 7×10 ft wind tunnel. Quantities compared include: surface pressures on the tractor/trailer, vehicle drag, and time-averaged velocities in the base region behind the trailer. The results indicate that both turbulence models are able to accurately capture the surface pressure on the vehicle, with the exception of the base region. The Menter k-ω model does a reasonable job of matching the experimental data for base pressure and velocities in the near wake, and thus gives an accurate prediction of the drag. The Spalart-Allmaras model significantly underpredicted the base pressure, thereby overpredicting the vehicle drag.

To better understand the flow mechanisms that contribute to the aerodynamic drag of heavy vehicles, unsteady largeeddy simulations are performed to model the wake of a truncated trailer geometry above a no-slip surface. The truncation of the heavy vehicle trailer is done to reduce the computational time needed to perform the simulations. Both unsteady and time-averaged results are presented from these simulations for two grids. A comparison of velocity fields with those obtained from a wind tunnel study demonstrate that there is a distinct difference in the separated wake of the experimental and computational results, perhaps indicating the influence of the geometry simplification, turbulence model, boundary conditions, or other aspects of the chosen numerical approach.

Results of drag reduction studies based on direct numerical simulations of the separating flow aft of a rectangular forebody are reported. The study examines the two-dimensional wake flow forming in the presence of rectangular boat-tails (blocks or plates) attached to the base of a forebody, with special emphasis on modifying the base pressure component of the drag. An optimal aspect ratio for boat-tail notches at the separation point are identified, and the underlying mechanism leading to elevated base pressures in such configurations is revealed.

The use of planar-sided boat tail plates for aft-end drag reduction on a tractor-trailer was studied numerically, experimentally and on a full scale prototype. Parametric wind tunnel tests utilized a 1:15 scale Peterbilt 379 tractor and 48 foot (14.6 m) trailer with cavity plate concepts mounted perpendicular to the trailer base. Yaw angles up to 9 degrees were examined. Qualitative numerical results confirmed a pressure increase on the aft face of the trailer. Model drag increments, obtained at zero yaw and a width-based Reynolds number of 230,000, based on trailer width, indicated reductions in the drag coefficient, based on frontal area, of up to 0.075 or about 9% of the baseline model trailer drag. Removal of the top plate degraded the performance of all devices. Performance also decreased with yaw angle for all plates mounted perpendicular to the trailer base, contrary to devices with angled plates. Devices with shorter angled plates indicated better performance with the top open rather than an open bottom. Drag reduction was more sensitive to plate inset from the trailer edge than to plate length and a zero inset of the bottom plate maximized performance. Two full scale prototypes were road tested, the first utilized rigid composite sides with a flexible top and bottom and the second was an all rigid-sided aluminum design. The former exhibited cross-country road fuel savings of about 0.5 miles per gallon (0.2 kilometers/liter), approximately 9%, over a 10,000 mile (16,093 km) trip, while the latter returned inconclusive results. Estimated fuel savings for a typical 120,000 miles (193,121 km) per year traveled were approximately 1500 gallons (5677 liters) per truck.

Numerical simulations of a truck-type configuration are carried out with the aim to investigate the effectiveness of various drag reduction devices. Unsteady Reynolds-Averaged Navier-Stokes equations are solved to capture the periodic wake motion at the base of the model and comparisons with available measurements show satisfactory agreement. Two passive devices are applied to the back of the model to investigate their effect on the pressure recovery and the base pressure. Up to 15% drag reduction is obtained using a boattail appendix. In addition, a Coanda jet system is designed and simulated. In this case the amount of drag reduction is directly related to the amount of injected mass and a zero drag condition can be obtained.

Under contract to the US Department of Energy, Georgia Tech Research Institute (GTRI) has developed and applied blown aircraft aerodynamic technology to entrain separated flowfields, significantly reduce drag, and increase the fuel economy of Heavy Vehicles. These aerodynamic improvements also lead to increases in stability and control, braking, and traction, thus enhancing safety of operation. GTRI wind-tunnel model results on test Heavy Vehicle (HV) models demonstrated drag coefficient reductions of 50% using only 1 psig blowing pressure in the plenums, and over 80% drag reductions if additional blowing air were available. Additionally, an increase in drag force for braking was produced by blowing different slots. Lift coefficient was increased for tire rolling resistance reduction, while down force could be produced for traction increase. Also, side force and yawing moment were generated on either side of the vehicle, and directional stability was restored by blowing the appropriate side slot. These experimental data confirmed the elimination of directional instability caused by side-winds.The above model data formed the basis for the design and modification of a full-scale test vehicle by prototype shop Novatek, Inc. and GTRI. Initial confirmation road test results are presented for this patented concept applied to an HV test rig supplied by team members Volvo Trucks of North America and Great Dane Trailers. To verify fuel economy increase, an SAE Type-II Fuel Economy test was conducted at the Transportation Research Center test track in East Liberty, Ohio. Results presented in this paper include wind-tunnel data for both unblown and blown configurations, full-scale blowing and fuel-economy data, comparisons to experimental results of the smaller-scale blown Pneumatic Heavy Vehicle model, and tunnel tests on a full-scale Pneumatic SUV.

The objective of this investigation is to study possible means for reducing the base drag of a tractor-trailer. The experiments are conducted in the Dryden wind tunnel at the USC Ground Vehicle Aerodynamics Laboratory. A roughly 1/15 scale model resembling a trailer is utilized for the study. The model is fitted with a shaped nose-piece to ensure attached flow over the forward portion of the model. The model is equipped with a force balance to measure drag. In addition base pressures are measured, and hot-wire wake surveys are conducted downstream from the model base. The Reynolds numbers (based on the square-root of the model cross-sectional area), range from 0.1 × 106 to 0.4 × 106.Drag reduction is effected by means of flaps attached along the edges of the model base, and inclined inward to decrease the size of the downstream wake. In addition, an oscillatory perturbation is applied at the flap origin in an attempt to maintain attached flow for larger angles of flap inclination.The present study has found that a simple, passive base-flap deflection—no forcing whatsoever—produces significant drag saving. The maximum drag reduction is 0.06 – 0.08 at an angle of 9–10 degrees. The magnitude of the saving is in accord with both early and recent measurements in other laboratories.The present results also show that oscillatory momentum addition has little effect on drag reduction unless the net oscillatory momentum flux coefficient is equal or greater than 0.1%. Increasing the oscillatory momentum perturbation to a coefficient value of 0.3% produces drag savings at angles greater than 9–10 degrees, but has very little effect upon the maximum saving at 9–10 degrees.To further study this anomalous behavior, follow-on experiments are planned to investigate a larger range of forcing amplitudes, and a variety forcing-function duty cycles. In addition, Digital Particle Image Velocimetry will be used to capture the detailed flow-field in the vicinity of the flap.

In the commercial vehicle department at DaimlerChrysler, CFD is extensively used for many different design tasks in the development of trucks and commercial vans. Since simulation software has become available allowing fast and easy setup of aerodynamic models, computational aerodynamics is playing an important role in earlier stages of the design development process.

This paper presents results from flow simulations about a simple train model to be used in an upcoming experiment. The simulations are made with the unstructured Navier-Stokes solver TAU employing hybrid grids. The influence of wind tunnel walls as well as a ground plate with and with out a boundary layer is examined. The results obtained so far indicate that the influence of the wind tunnel wall can be neglected in the computation, while the boundary layer on the ground plate might become critical in the experiment. Furthermore the influence of cross wind on the train is studied by yawing the train up to 30° to the oncoming flow.

Turbulence model development for aerodynamic applications has for many years concentrated on improving the capabilities of CFD methods for separation prediction. Validation studies of turbulence models in the ‘80th have clearly shown that most turbulence models were not capable of predicting the development of turbulent boundary layers under adverse pressure gradient conditions. Based on that observation, new models were developed to specifically meet this challenge, resulting in a series of models capable of capturing boundary layer separation in good agreement with experimental data (Johnson and King 1984, Menter 1993, Spalart and Allmaras 1994).

Recent developments of the 3-D Lagrangian vortex element method for bluff body flows are presented. In this approach attached boundary layer regions are modelled using infinitely thin vortex sheets while Lagrangian vortex elements are used for the separation regions and the wake. Preliminary results for the flow past a simplified generic truck geometry are presented. Further developments, aimed at the development of a hybrid Eulerian-Lagrangian solver, are briefly introduced.

An up front CAE coupling of tools is described which provides for routine thermal flow analysis of vehicle powertrain cooling systems for a variety of vehicle shapes and sizes. A family of parametric models is generated using Pro/E which allows for flexible representation of a variety of vehicle classes with accurate thermal flow analyses. The parametric CAD models are morphed to model the important flow features in a relatively complex underhood vehicle model. Some detailed modeling is often added to the generic parametric model for flexibility in modeling important unique details. All tetrahedra meshes are generated automatically from the CAD model using Simmetrix software. High aspect ratio elements are layered near appropriate surfaces for appropriate modeling of shear layers. Templates are used for exporting the mesh and flow simulation parameters to the flow solver, AcuSolve. An unsteady, incompressible air flow with energy equation and Spalart-Almaras turbulence model is used underhood and around the vehicle with simple heat exchanger models and coupled radiation. The flow through grille opening elements is modeled with detailed surfaces and appropriate grid resolution of flow. An adaptation technique using an a posteriori approach is demonstrated. Comparison to test data shows good correlation for a variety of steady and unsteady cases.

The management of underhood temperatures is becoming ever more important as emissions and noise regulations become more severe. The emissions regulations typically add to the heat load, while the noise regulations lead to a more tightly sealed engine compartment.

To gain insight into the ventilation needs for an enclosed engine compartment of an off-road machine, a prototypical test-rig that includes an engine and other installation hardware was built. Well controlled experiments were conducted to help understand the effects of ventilation air flow on heat rejection and component temperatures. An assessment of 1-D and 3-D simulation methods was performed to predict underhood ventilation air flow and component temperatures using the experimental data. The analytical work involved development, validation, and application of these methods for optimized ventilation air flow rate in the test-rig. A 1-D thermal-fluid network model was developed to account for overall energy balance and to simulate ventilation and hydraulic system response. This model was combined with a 3-D CFD model for the ventilation air circulation in the test rig to determine the flow patterns and the distributed surface heat transfer. The tests conducted at Caterpillar and the complementary analyses performed at Argonne provide an opportunity to understand the isolated effect of ventilation air cooling on underhood thermal management.

Computational Fluid Dynamics (CFD) analyses were performed for an armored tank engine compartment cooling flow. Large hybrid unstructured meshes (2.5–3.0×106 cells) were constructed using the ICEM-CFD grid generator. The flow and convective heat transfer field were computed using an in-house CFD code, NPHASE. The commercial software package, RADTHERM was utilized to incorporate radiation heat transfer within the simulations.Two steady operating conditions and one engine-off cool-down transient were analyzed. Specifically, the conditions analyzed were open throttle (hereafter OT), Tac Idle (TI) and engine off soakback (SB). OT and TI were run with and without convection heat transfer employed in the radiation assessments to provide best-estimate and conservative peak temperature predictions respectively. SB was run transiently using fixed heat transfer coefficients obtained from NPHASE analysis.Results are presented for the simulations performed, with emphasis placed on peak temperatures of several design critical elements.

The advent of the modern high-speed train system may be dated back to 1964 with the opening of the first Shinkansen high-speed line in Japan, at an initial revenue service speed of 210 km/h. At same time it associated very fast rolling-stock and dedicated high-speed track infrastructure, it opened a new era for the aerodynamic design of ground transportation systems. On the one hand, the aerodynamic portion of the train running resistance was going over two third of the global running resistance. On the other, new kind of aerodynamic problems needed to be faced, such as those going together with train travels in tunnel and passing of opposite high-speed trains on adjacent tracks.

When a train travels in the open air it displaces the air around and over it forming a slipstream alongside the train and a wake behind. The air at the surface of the train moves at the speed of the train, whilst far from the train the air moves at the ambient air speed. Therefore, there is a region near the train sides where the air can be moving at speeds comparable to that of the train. In this region the air is very turbulent and, depending on the aerodynamic roughness of the train, may contain complex and interacting vortices. After the train passes there is a wake flow, which decays as the train moves away.

The aerodynamic effects occurring in a tunnel as a train moves into or through it are totally different from those observed in the open air and their amplitude and severity grow as the train speed is increased. The flow in the whole tunnel needs to be considered in the same time as the flow in the vicinity of the vehicle. Aerodynamic forces, pressure waves and acoustics have a strong impact on safety and comfort issues.When a train enters into a tunnel, a compression wave is generated, propagates through the tunnel and is reflected at the tunnel extremity. During the reflection process a part of the wave is transmitted outside the tunnel in the form of a micro-pressure wave, which may generate a “sonic boom” problem, depending on the shape of the incident wave, in particular the gradient of the wavefront. The shape of the wave changes as it propagates through the tunnel under the influence of the unsteady viscous effects (in particular skin friction at the tunnel wall), the non-linear effects and the presence of material and components in the tunnel (for example, ballast or niches). Measurements of the skin friction behind a pressure wave are presented.

The lateral vibration of high-speed trains in tunnels has recently become a subject of discussions concerning riding comfort. The paper describes the phenomenon, its mechanism and countermeasures.First, running tests revealed that the aerodynamic force in tunnel sections is much greater than that in open sections. The aerodynamic force and the vibration in tunnel sections gradually increased from the head toward the tail of a train set; and the yawing vibration of cars had a close relation with the aerodynamic force.Second, to clarify the interaction between the vehicle dynamics and the aerodynamic force, the flow field around a scale model, which was forcibly vibrated, was analyzed by a wind tunnel experiment. The results showed that a pressure field that had the same properties as those of real trains was found even though the train model did not vibrate. The effect of vibration on the flow field was small and thus the phenomenon was considered as a forced vibration by the aerodynamic force.Third, to investigate the aerodynamic force, numerical simulations were conducted. The computation proved that the cause of the large pressure fluctuation at the tail is the flow separation by the sudden expansion of the effective flow area. It also revealed that the flow becomes unstable under the train. The resulting vortices are spread on the train side by the tunnel wall, and then the unsteady aerodynamic force is generated when the vortices pass.Finally, to derive an optimal shape, which suppresses the unsteady aerodynamic force, scale model tests were conducted. The results showed that a long nose effectively decreases the large pressure fluctuation at the tail. Rounding the lower section of the car and installing fins under the train were also shown to be effective countermeasures for reducing the unsteady aerodynamic force.

Although trains may be considered the safest existing ground vehicles, there has always been the occasional derailment due to strong side winds, mostly on islands such as Japan or the British Isles [1]. According to meteorologists, the current global warming leads to ever stronger winds meaning such accidents are more likely to happen not only in coastal areas but even in the very heart of larger continents. This is how a commuter train was blown over in Austria recently (Figure 1), only 5 years after a similar accident happened in Belgium.

This paper deals with numerical studies in relation with cross-wind effects on high-speed trains. 3-dimensional steady RANS simulations were performed for the ALSTOM Transport company with the off-the-shelf STAR-CD software. The purpose of the study was to investigate the aerodynamic performances of three different designs of very high-speed trainsets with airflow yaw angles ranging from 0 to 90 degrees. Each trainset was composed of five simplified vehicles resting over a flat ground.

The issue of energy economy in transportation has grown beyond traditional concerns over environment, safety and health to include new concerns over national security and energy self-sufficiency. As part of the U.S. Department of Energy Office of FreedomCAR and Vehicle Technologies’ Working Group on Aerodynamic Drag of Heavy Vehicles, Argonne National Laboratory is independently investigating the accuracy of aerodynamic drag predictions generated by commercial Computational Fluid Dynamics (CFD) Software. In this validation study, computational predictions from two commercial CFD codes, Star-CD [1] and PowerFLOW [2], will be compared with detailed velocity, pressure and force balance data from experiments completed in the 7 ft. by 10 ft. wind tunnel at NASA Ames [3,4] using a Generic Conventional Model (GCM) that is representative of typical current-generation tractor-trailer geometries. This paper highlights results from evaluations of drag coefficient predictions using standard two-equation steady RANS turbulence models and logarithmic wall functions that were completed as part of the first phase of these studies.

Aerodynamic characteristics of a ground vehicle affect vehicle operation in many ways. Aerodynamic drag, lift and side forces have influence on fuel efficiency, vehicle top speed and acceleration performance. In addition, engine cooling, air conditioning, wind noise, visibility, stability and crosswind sensitivity are some other tasks for vehicle aerodynamics All of these areas benefit from drag reduction and changing the lift force in favor of the operating conditions. This can be achieved by optimization of external body geometry and flow modification devices. Considering the latter, a thorough understanding of the airflow is a prerequisite.The present study aims to simulate the external flow field around a ground vehicle using a computational method. The model and the method are selected to be three dimensional and time-dependent. The Reynolds-averaged Navier Stokes equations are solved using a finite volume method. The Renormalization Group (RNG) k-ε model was elected for closure of the turbulent quantities.The external aerodynamics of a heavy truck is simulated using a validated computational fluid dynamics method, and the external flow is presented by computer visualization. Then, to help the estimation of the error due to two commonly practiced engineering simplifications, a parametric study on the tires and the moving ground effect are conducted on full-scale tractor-trailer configuration. Force and pressure coefficients and velocity distribution around tractor-trailer assembly are computed for each case and the results compared with each other.

This paper describes recent works of practical applications of vortex element methods to study of aerodynamics of heavy vehicles, carried by the authors’ group, explaining the mathematical basis of the method based on the Biot-Savart law. It is pointed as one of the most attractive features of the vortex method that the numerical simulation using the method is considered to be a new and simple technique of large eddy simulation, because they consist of simple algorithm based on physics of flow and it provides a completely grid-free Lagrangian calculation. As typical results of aerodynamics of heavy vehicles, unsteady flows around a heavy vehicle model such as a tractor-trailer with different gap lengths and unsteady aerodynamic characteristics of a tractor-trailer with meandering motion are explained.

The development of a suitable mounting system to test a heavy duty truck model in a wind tunnel is an evolutionary process. The mounting system is usually designed at the beginning of a test program based on the wind tunnel current layout, past experiences and the initial objectives of the program. If the test program becomes a regular event, periodic modifications to the mounting system may become necessary in order to refine and reduce its tare and interference. International and the Oran W. Nicks Low Speed Wind Tunnel at Texas A&M have more than 20 year history in the development of such a mounting system. The history starts with a single strut mount system with a ground board follows through a change to a two strut system without the ground board, a change to a four-point mounting system and ends with a six-point system. The development also required modifications to the wind tunnel test section, which helped increased model strength with minimum changes. The results clearly show significant reduction in tare and interference of the six-point mounting system over its predecessors.

Existing models suggest base drag is dependent upon forebody drag, however, these models do not provide accurate predictions when applied to large-scale vehicles. This paper describes preliminary investigations into a new base drag model and the feasibility of minimizing total drag by optimizing the forebodydrag to base-drag relationship.

This paper presents results of a series of road tests measuring the spray clouds generated by a truck/trailer traveling on a watered roadway with and without specifically-configured splash guards covering all the wheels. The tests were conducted with a carefully-arranged set of laser transmissometers, and data show significant reductions in spray density at some fixed positions from the truck-trailer and the roadway. Data is also given showing the effect of adding a drag-reducing air shield on the truck. Emphasis is put on the full-scale test procedures and on the reliability/repeatability of the measurements.

GTRI has recently been developing pneumatic aerodynamic concepts for application to Heavy Vehicles under a Department of Energy contract through the Oak Ridge National Laboratory (ORNL). A related application under development is a novel heat exchanger known as the Aerodynamic Heat Exchanger (AHE). This patented device employs airfoil/wing aerodynamic pressure differences to induce large mass flows across a radiator installed inside a wing. GTRI has recently completed an in-house wind tunnel test of this concept. The objective of this proposed effort was to perform a wind-tunnel evaluation of the AHE and establish the feasibility of the concept. A 2D wing was fabricated with a removable center section. A radiator core was integrated into this section of the wing. A conventional radiator core (based on a Visteon design) and two cores made from carbon foam were tested. The carbon foam cores were designed and provided to GTRI by ORNL. Hot water was allowed to pass through the inside of the wing while freestream, wind tunnel air passed over (and through) the wing. Heat rejected by the radiator was measured as well as lift and drag. Results indicated that the concept is feasible and can provide an effective means to reduce vehicle drag by reducing the drag due to conventional radiators.

Automation technology can expand the performance envelope for heavy vehicles by eliminating the limitations imposed by driver performance. Accurate automatic steering control makes it possible for the vehicles to operate safely, and over a full speed range, within lanes only slightly wider than the vehicles themselves.