WO2019155181A1

WO2019155181A1 - Drag reduction

Info

Publication number: WO2019155181A1
Application number: PCT/GB2018/050348
Authority: WO
Inventors: Jonathan MORRISON; Juan Marcos GARCIA DE LA CRUZ; Georgios RIGAS; Andrew WYNN; Rowan BRACKSTON
Original assignee: Imperial College Of Science, Technology And Medicine
Priority date: 2018-02-07
Filing date: 2018-02-07
Publication date: 2019-08-15

Abstract

A drag reduction apparatus for reducing the aerodynamic drag on a bluff body with a blunt trailing edge caused by fluid flow characteristics at the wake of the bluff body, the drag reduction apparatus comprising one or more control elements configured to be coupled to the bluff body and to be positioned asymmetrically relative to a direction of travel of the bluff body in response to a determined parameter of the bluff body to control fluid flow at the wake of the bluff body and reduce the drag experienced by the body.

Description

DRAG REDUCTION

Field The present disclosure relates to a drag reduction method and a drag reduction apparatus.

Background Many vehicles, usually road vehicles, are designed in a particularly aerodynamically inefficient fashion: their shape often contains a blunt-edged rear end, which is the result of legal and operational but also aesthetic constraints. This is the case for a number of different vehicle types, for example heavy-duty vehicles, trucks, large road vehicles (LRV), vans, minivans, compact cars, city cars, SUV, MPV, station wagons or off-roaders.

From an aerodynamic perspective a blunt end of a“bluff” body, such as a vehicle, is highly inefficient: it generates a wake, which can dissipate as much as 25% of the total energy produced by the engine. Improving the aerodynamic efficiency in vehicles is paramount for environmental, economic and practical reasons. Also, for the emerging industry of electric cars, an improvement in aerodynamic efficiency is related with an increment in range, helping to overcome“range anxiety”. The relevance of improving the aerodynamic efficiency of vehicles is now universally accepted and many companies and brands provide different solutions. It is gradually becoming more common to see Large Road Vehicles using fairings, diffusers or boat tails but also, most modern car manufacturers such as Audi, Porsche, Volkswagen or Renault include in their designs relatively large spoilers (up to 20% of the vehicle height) to control the vehicle wake.

It is the intention of this disclosure to contribute to improving the aerodynamic efficiency of vehicles using a technology characterised by its small geometric impact, in comparison with other technologies currently present in the market. In

l particular, the technology can be regarded as a small, but crucial, modification of the spoilers currently used in some vehicles, so it will effectively not add any extra geometric feature to the current designs. In other applications, such as for vehicles which do not use spoilers, the disclosure can provide improved aerodynamic efficiency by the addition of a drag reducing apparatus to a bluff body.

Summary A drag reduction apparatus for reducing the aerodynamic drag on a bluff body with a blunt trailing edge caused by fluid flow characteristics at the wake of the bluff body, the drag reduction apparatus comprising one or more control elements configured to be coupled to the bluff body and to be positioned asymmetrically relative to a direction of travel of the bluff body in response to a determined parameter of the bluff body to control fluid flow at the wake of the bluff body and reduce the drag experienced by the body.

Optionally, the drag reduction apparatus further comprises one or more actuators configured to cause the one or more control surfaces to move with respect to the bluff body. Optionally, the drag reduction apparatus further comprises a processor configured to execute a control algorithm so as to generate at least one control signal for controlling the position of the one or more control surfaces. Optionally, the position of the one or more control surfaces is a rotational position. Optionally, the processor is configured generate the control signal based on the parameter of the bluff body. Optionally, the drag reduction apparatus further comprises at least one sensor configured to determine an environmental parameter, and the processor is configured to determine the parameter of the bluff body based upon an output of the at least one sensor. Optionally, the environmental parameter is the stagnation pressure at one or more locations towards the front of the bluff body. Optionally, the at least one sensor is at least one pitot tube. Optionally, the processor is configured to apply a low pass filter to the output of the sensor.

Optionally, the processor is configured to determine the effective yaw angle of the bluff body based upon the determined stagnation pressure. Optionally, the processor is configured to generate the control signal based on the determined effective yaw angle and a lookup table of effective yaw angles and control surface positions.

Optionally, the parameter off the bluff body is the drag currently experienced by the bluff body. Optionally, the processor is configured to determine the drag based on the current pressure at the rear of the body. Optionally, the processor is configured to determine the drag based on a force currently experienced by the one or more control surfaces. Optionally, the bluff body is driven by an engine and the processor is configured to determine the drag based on the current engine power.

Optionally, the bluff body is a road vehicle. Optionally, the road vehicle is a heavy goods vehicle“HGV” or a light goods vehicle“LGV”. Also disclosed in a vehicle comprising the drag reduction apparatus. Optionally, the vehicle comprises a plurality of pressure sensors positioned on the vehicle so as to measure pressure in the wake of the vehicle. Optionally, the vehicle comprises a plurality of control surfaces, each coupled to the vehicle at a rear of the vehicle so as to control the fluid flow at the wake of the vehicle. Also disclosed is a drag reduction method for reducing the aerodynamic drag on a bluff body with a blunt trailing edge caused by fluid flow characteristics at the wake of the bluff body, the drag reduction method comprising coupling one or more control elements to the bluff body, determining a parameter of the bluff body, and positioning the one or more control elements asymmetrically relative to a direction of travel of the bluff body in response to the sensed parameter to control fluid flow at the wake of the bluff body and reduce the drag experienced by the body.

Optionally, the method further comprises moving the one more control surfaces by moving one or more actuators configured to cause the one more or more control surfaces to move to control fluid flow. Optionally, the method further comprises executing a control algorithm using a processor to generate at least one control signal for controlling the one or more actuators. Optionally, the method further comprises controlling a rotational position of the one or more control elements. Optionally, the method further comprises generating the control signal based on the parameter of the bluff body. Optionally, the method further comprises determining the parameter of the bluff body by detecting an environmental parameter using at least one sensor. Optionally, the environmental parameter is the stagnation pressure at one or more locations towards the front of the bluff body. Optionally, the at least one sensor is at least one pitot tube. Optionally, the method further comprises applying a low pass filter to the output of the sensor. Optionally, the method further comprises determining the effective yaw angle of the bluff body based upon the determined stagnation pressure. Optionally, the method further comprises generating the control signal based on the determined effective yaw angle and a lookup table of effective yaw angles and control surface positions. Optionally, the parameter off the bluff body is the drag currently experienced by the bluff body. Optionally, the method further comprises determining the drag based on the current pressure at the rear of the body. Optionally, the method further comprises determining the drag based on a force currently experienced by the one or more control surfaces. Optionally, the method further comprises driving the bluff body with an engine and determining the drag based on the current engine power.

Optionally, the bluff body is a road vehicle. Optionally, the road vehicle is a heavy goods vehicle“HGV” or a light goods vehicle“LGV”.

Also disclosed is a computer-readable medium comprising computer-readable instructions which, when executed by a processor, cause the processor to perform the method. At high speeds of fluid flow the wake of bluff bodies, such as vehicles, contributes to a large percentage of their aerodynamic drag and total energy consumption.

Wakes are highly dissipative regions behind bodies moving within a fluid, characterised by high turbulence intensity and large coherent flow structures. The turbulence in wakes is sustained by several flow instability mechanisms that produce quasi-periodic or random oscillatory patterns in the fluid. Moreover, flow asymmetries, produced by the flow instabilities and by environmental asymmetries, increase the energy dissipation and total entropy of the wake. Among these instabilities, the most widely known are vortex shedding, bistability, and shear layer modes.

By providing the claimed features, the effect of environmental asymmetries can be reduced. As a result, the energy dissipated by the wake and the resulting aerodynamic drag is reduced. To achieve this, one or more control elements in the form of surfaces may be located at different positions on a bluff body, for example at the rear of the body with respect to the usual direction of travel of that body. These control surfaces can be considered as dynamic because, unlike static spoilers, they are configured to move in position with respect to the body during use to reduce wake intensity and drag. In some arrangements, the control surfaces can be operated in open loop (without sensors), in closed loop (with sensors distributed around the body) or in a combination of open and closed loop.

Brief description of the drawings Embodiments will be described below, by way of example only, with reference to the accompanying drawings, in which:

Figure 1 is an isometric view of a prior art bluff body; Figure 2a and 2b are top views illustrating different features of the wake generated by the bluff body of Figure 1 ;

Figure 3 is an isometric view of an illustrative drag reduction apparatus coupled to a bluff body according to an example of the disclosure;

Figure 4 illustrates a schematic diagram of an implementation of the disclosed system; Figure 5 illustrates the effect on drag of yaw angle of the bluff body of Figure 1 and control surface deflection angle on drag for a given yaw angle;

Figure 6 illustrates the average position and standard deviation of the base centre of pressure at a yaw angle of 2.5° at different mean control surface offsets or deflections;

Figure 7a shows a probability density function of wind speed at Fleathrow airport; Figure 7b shows the effects of yaw angle at two highway speeds associated with the distribution shown in Figure 7a;

Figure 8 shows an experimental set-up including an Ahmed body of width W 216mm;

Figure 9a illustrates the drag change of a bluff body having various symmetric arrangements of control surfaces optimised for given yaw angles and by a bluff body having adaptive control of the position of the control surfaces, the control surfaces having an extent of 9% of the vehicle width (Figure 8) in comparison to a baseline (no control surfaces) at a zero-yaw angle;

Figure 9b illustrates the data of Figure 9a in terms of the drag reduction achieved at different yaw angles by the optimised arrangements of control surfaces in comparison to a body having no control surfaces;

Figure 10a illustrates the drag change of a bluff body having various symmetric arrangements of control surfaces optimised for given yaw angles and by a bluff body having adaptive control of the position of the control surfaces, the control surfaces having an extent of 13% of the vehicle width in comparison to a baseline having no control surfaces at a zero-yaw angle;

Figure 10b illustrates the data of Figure 10a in terms of the drag reduction achieved at different yaw angles by the optimised arrangements of control surfaces in comparison to a body having no control surfaces; Figure 11 shows the average drag saving percentage of five different flap positioning strategies weighted with yaw angle distribution obtained for Heathrow at vehicle speed 60mph or 70mph and flap lengths 9% and 13% of vehicle width;

Figure 12 shows contour maps of the drag coefficient for flap lengths 9% and 13% of vehicle width, flap angles between -5° and 25° and yaw angles of 0°, 3°, 6° and 9°; Figure 13 shows an example arrangement of sensor locations on a vehicle;

Figure 14 shows a control algorithm for a drag reduction apparatus according to an embodiment; and Figure 15 shows another control algorithm for a drag reduction apparatus according to an embodiment.

Detailed description The following embodiments relate generally to a drag reduction apparatus and method. The application disclosed herein refers to front and rear surfaces of the bluff body. These terms are intended to refer to the nominal“front” and“rear” of the body with respect to the general flow of fluid over the body. Typically, the flow of fluid over the body will be from the front of the body to the rear. However, it will be appreciated that the direction of fluid flow may vary with respect to the body. Therefore, the term“rear” is intended to refer to the end of the object where the wake is formed due to fluid flow.

Where the bluff body is a vehicle, the“front” side or end of the body refers to the end of the vehicle that faces the direction of travel during forward travel of the vehicle and the“rear” side or end refers to the end of the vehicle opposing the direction of travel during forward travel. An isometric view of a prior art bluff body 100 having a blunt surface 110 is illustrated in Figure 1. Fluid flow over the body 100 results in a wake that has characteristics that contribute significantly to the drag on the body 100. The principle by which the disclosed approach reduces drag in bluff bodies is by minimizing the kinetic energy transferred from the body to its wake, either due to quasi-periodic fluctuations, triggered by self-excited fluid structures, as illustrated in Figure 2a; or to stable base pressure asymmetries, which favour large scale fluid recirculation, therefore seeding vorticity to the wake, as illustrated in Figure 2b. Different features of the wake generated by the bluff body of Figure 1 may occur simultaneously but are illustrated separately in Figures 2a and 2b. In particular, Figure 2a exemplifies vortex shedding and Figure 2b exemplifies bistability.

An example drag reduction apparatus of the present disclosure is illustrated in Figure 3 attached to a bluff body 300 near the blunt surface of the body. The drag reduction apparatus of Figure 3 comprises one or more control surfaces 320 positioned on a blunt surface 310 of the bluff body 300 towards the rear end of the body. The control surfaces 320 are flap-like surfaces each coupled to the body 300 and configured to move with respect to the body 300, for example with pivotal movement about a coupling point. Each control surface is a flap-like surface that may be elongated and may be able to direct fluid flow across its surface. In some arrangements, each control surface may be located at the rear of the bluff body. For example, each control surface may be positioned on the side of a rear surface of the body or on the side of the body towards the rear of the body. The control surfaces 320 may be moved independently. The control surfaces 320 of Figure 3 are each also coupled to an actuator 330. Each actuator 330 is configured to cause the coupled control surface 320 to move with respect to the body 300 based upon a control signal received by the actuator 330 from a control unit (not shown). More detail relating to the actuators 330 is given in patent document WO 2017/072530.

In addition, the drag reduction apparatus in the example of Figure 3 is configured to comprise one or more sensors 340. The sensors 340 may be pressure sensors which are located on the rear surface of the body 300. The pressure sensors may be configured to obtain pressure measurements at the rear of the body 300 to estimate the state of the wake. Different arrangements of control surfaces 320 and sensors 340 with respect to the rear of the body 300 are possible. Different types of sensor can be implemented at different locations anywhere on the body as appropriate, as will be explained.

Figure 4 is a schematic diagram of an implementation of the disclosed system 400. The system comprises a vehicle 402 and a drag reduction apparatus 404 such as that shown in Figure 3. The vehicle 402 comprises a power unit 406, a battery 408, an electronic control unit (ECU) 410 and a chassis/coach 412. The drag reduction apparatus comprises a control unit 414, sensors 416, such as the sensors 340, control surfaces 418, such as the control surfaces 320, and actuators 420 to move the control surfaces 418, such as the actuators 330. The power unit 406 of the vehicle 402 provides power to the battery 408 and the ECU 410. When the drag reduction apparatus 404 is installed on the vehicle 402, the battery 408 supplies power to the control unit 414 , the sensors 416 and the actuators 420. The sensors 416 may require less power than the other components. Alternatively, the drag reduction apparatus 404 may comprise a dedicated power supply for controlling the control unit 414 and the actuators 420. In operation, the sensors 416 provide sensor data to the control unit 414. The control unit 414 processes this data to provide a control signal to the actuators 420. In some embodiments, the control unit 414 runs a control algorithm with an input signal from at least one sensor 416 configured to determine at least one environmental parameter, and then generates a control signal based on that output. The control signal may control the position of one or more of the control surfaces 418. The control signal may also control oscillation of one or more of the control surfaces 418 about a mean position, in the manner described in WO 2017/072530. The actuators 420 in turn move the control surfaces 418 to the position indicated by the control signal.

Alternatively or additionally, the control unit 414 may receive information from the

ECU 410, such as velocity, fuel consumption, steering wheel position, motor rpm, which is received by the ECU 410 from sensors that may already be installed by default in the vehicle telemetry system. The control unit 414 could be implemented as field-programmable gate array (FPGA) - programmable logic blocks which then operate as“black boxes” and can be updated in subsequent versions. Various categories of sensors may be used for data acquisition. Base- flap contained sensors can also be installed within a base-flap, such as control surface 320. These sensors do not require communication with other parts of the vehicle as they would be embedded as part of the flap. This could be appealing as a simple retrofitting option. These sensors would detect variables such as flap deflection, moment and forces on flaps, 3-axis acceleration, GPS, etc.

In some arrangements, a processor in the control unit 414 executing a control algorithm may receive pressure measurements obtained by a plurality of pressure sensors. The control algorithm may perform spatial averaging of pressure measurements from the pressure sensors. The flow features to be controlled are large scale but coexist alongside multiscale turbulence. It is therefore preferable to extract features of interest from noisy measurements. Averaged pressure statistics such as the Centre of Pressure location (CoP) or spatial mode projection give a single metric from multiple pressure measurements, thereby extracting relevant information.

In some arrangements, the control algorithm aims to achieve a desired value for the aforementioned average pressure metric. For the case of bi-stability control with zero cross-wind, it is desirable to maintain the CoP position at the centre of the base in order to achieve a symmetric wake. Under other conditions the optimal target value may be determined in real time based on knowledge of the flow conditions, or by using an extremum seeking algorithm.

In order to set the pressure metric to a target value, a negative feedback gain may be used in the control algorithm to drive the position of the control surfaces in the appropriate direction. In addition to this, a combination of frequency domain filters and compensators may be used in order to improve stability and suppress unwanted oscillations. The specific nature of the spatial averaging and controller may be determined based upon a combination of mathematical models for the wake and actuator, as well as empirical results from experiments. The system described in patent document WO 2017/072530 is well-adapted to front-on flow, with a zero yaw angle. However, it is typical for bluff bodies, such as road vehicles, to operate at significant yaw angles with respect to the flow of fluid (where fluid approaches the body at an angle), which drastically increases the drag experienced. This is mainly due to the typical cross-winds experienced by these vehicles. The yaw angle, b, may be defined as the relative angle of airflow to a direction of travel of the bluff body. As an example, a crosswind at the average wind speed in the UK, about 5m/s, would produce a yaw angle of 10.6^° in a truck driving at 60mph. This will increase the vehicle drag by more than 20%, as presented in Figure 5a. In addition, head winds modify the velocity profile so that the body faces away from the uniform expected for the zero-wind baseline and change their pitch, i.e. their vertical yaw angle. Environmental asymmetries may encompass any geometrical vehicle-flow misalignment that, in a steady state, lead to an asymmetric optimal position of the aerodynamic surfaces for drag reduction. Other environmental asymmetries to be considered in the example of vehicles are road turns or vehicle operational asymmetries such as wakes induced by other upstream vehicles, open windows, protrusive objects, load asymmetries or atmospheric boundary layer variations.

The inventors have recognised that modifying the mean position of the control surfaces 320 with respect to the body 300, for example by applying and/or adjusting an offset to a control signal that controls the position and movement of the control surfaces, can recover some of the base pressure lost due to these environmental asymmetries, thereby reducing the drag that they generate, as shown in Figure 5b. The combined effect of environmental asymmetries often varies over long time scales. Minimizing the resultant drag may involve an adaptive algorithm, which sets independently the position of each control surface 120, including the mean position and/or the oscillation frequency of the control surface 120, as a function of the real-time data acquired. Figure 5a presents the position of the base centre of pressure at a yaw angle of 2.5° and different mean control surface offsets or deflections. It is noteworthy that, as observed comparing Figure 5b and Figure 6a, in the presence of environmental asymmetries the maximum drag reduction is not necessarily obtained by increasing the base pressure distribution symmetry.

Among the stable base pressure asymmetries, the inventors have identified two main sources of drag: environmental fluctuations and wake bistability. The quasi- periodic wake oscillations targeted in the drag reduction apparatus are the vertical and horizontal vortex shedding. The inventors have recognised that bi-stability can also occur in non-symmetric configurations, such as in the case of cross- wind or any other environmental asymmetries. In that case, the bi-stable solution will, once more, restrict the maximum drag reduction attained. Figure 6b illustrates the appearance of the bi-stability in an asymmetric situation as the increment of the standard deviation in the lateral centre of pressure position at flap deflections over 10°. Reducing the drag further may require the use of a closed-loop system which positions the wake in a naturally unstable solution, just as in the case of the symmetric configuration.

The standard vehicle operation happens within a windy environment. As an example, Figure 7a presents the probability density function of the wind speed at Fleathrow airport. Other areas of the UK, EU and USA present wind distributions similar enough to consider the case presented of general relevance.

Figure 7b presents the probability distribution of the yaw angle of a vehicle driving at two speeds (60 and 70 mph) within an environment characterized by the wind distribution presented in Figure 7a considered uniformly distributed over all directions. Even if the wind distribution in general is not asymmetric, the angle range of the wind rose and the approximately uniform distribution of road directions make this assumption fair. As observed, the distribution of yaw angle is almost uniform up to 6° and relevant up to 12°. In this analysis, the range 0° to 9° is considered, which, according to the figure, happens around 80% of the time. To study the effect of such environmental conditions, an Ahmed body 800 (general road vehicle model used in aerodynamic experiments) as illustrated in Figure 8 was tested in a wind tunnel at 35 m/s without flaps (baseline) and with rear lateral flaps/control surfaces 810. In the experimental set up employed, two flaps are located at the rear lateral edges of the body 800 and respectively set at angles qi and Q2 with respect to the model. To evaluate the efficacy of different flap positioning strategies under cross-wind, the model was tested in a wind tunnel, with the Reynolds number, Rew, of the order 10⁵, with and without flaps at yaw angles b = 0°, 3°, 6° or 9°. The flap sizes tested, d, were 9% or 13% of the body width.

Results of this experiment are shown in Figures 8 to 17. Figures 8a and 9a present the drag of the (no-flap) baseline and the drag of the body using different flap/control surface 810 positioning at the four yaw angles tested, as a percentage of the baseline at b=0°, for flap lengths 9% and 13% of the body width respectively. The flap 810 configurations selected in each case are those symmetric ( Q₁ = (¼) flap positions that maximize the drag reduction at the yaw angles tested, and the flap positioning (in general asymmetric, Q₁ ¹ (¼) that maximizes the drag reduction at each particular yaw angle. This last optimal flap positioning is different at each yaw angle.

Figure 9b and Figure 10b present the same results as Figure 9a and Figure 10a as the drag saving with respect to the no-flap baseline at each yaw angle. As observed, except for the configuration aligned with the flow, the asymmetric flap configuration consistently produces a relevant increase in drag reduction with respect to each optimal symmetric flap configuration.

Moreover, a non-adaptive system cannot optimize the symmetric configuration of the flaps 810 at each yaw angle. That is to say, any given static optimum angle identified will only be optimal for a particular yaw angle. As such, a more realistic comparison for an adaptive system against a static system is to look at the average of savings achieved weighted with the yaw distribution presented in

Figure 7b. This is illustrated in Figure 11. In this figure, a weighted average of the different flap positioning strategies for the vehicle speeds 60mph and 70mph and the two flap lengths considered are presented. As observed, the adaptive solution provides a drag reduction 40% to 70% larger than the best symmetric static positioning strategy, depending on the flap length and yaw angle distribution expected. This data clearly indicates the benefits of adaptive systems. These can react to variations in environmental conditions and can provide significantly larger drag reductions than static systems such as the boat-tails currently commercialized, for typical wind distributions and vehicle velocities.

It has been discussed above that, for non-zero yaw conditions, asymmetric flap configurations produce a relevant increase in drag reduction. For each b and d the contour maps of the drag coefficient, Co, against qi and Q2 are presented in Figure 12. The global minimum of these CD surfaces, marked with a x, is generally not located under a condition of symmetric flap deflection, Q₁ = 0₂, marked with a line. Any static positioning necessarily complies with such condition, making it equally effective for positive and negative b. The minimum C_D across such symmetric conditions is marked with a +. As observed, an increase in b leads to a more deflected symmetric configuration to optimize the drag reduction. When asymmetric flap deflection is allowed, the outside flap angle, Q₂, increases by a factor of up to about 50%, whereas the inside flap angle, Q₁ is reduced by a factor of about 100%, even becoming negative (i.e. deflecting outwards) in some cases. Increasing the flap length, d, is expected to provide the flaps with higher authority. This leads to larger symmetric deflections achieving the minimum CD as well as larger optimal values for both Q₁ and Q₂ at the global minimum. The curvature of the C_D surface in this absolute minimum is also seen to be larger, and hence the gradient of the C_D surface around this minimum, for longer flaps. Therefore, C_D reductions benefit further from a more precise flap deflection for longer flaps.

Under zero yaw angle conditions, the optimal flap deflection is seen to be both symmetric and inwards, consisting of a pure boat-tailing effect, consistent with the understanding that a narrow and symmetric wake corresponds to the lowest drag. Further to this it is seen that for flap angles 0 < 15°, the optimal condition is always symmetric, demonstrating that any asymmetry imposed by the flaps is detrimental to the drag. Flowever, under conditions for which b ¹ 0°, the maps instead exhibit approximate symmetry about a Q₂ = constant axis, meaning that regardless of Q₁ the optimal value for Q₂ is approximately constant for a given condition. Similarly, for a fixed value of Q₂ and b ¹ 0°, the optimal value of Q1 is approximately constant, lying along a vertical line at Q₁ ~ 0°. In order to understand the factors that control the optimal flap angles, it is insightful to also look at other measurements such as the base suction, -<C_P> quantifying the pressure drag, and the lateral force coefficient CL Such data for one set of conditions is shown in Figure 12, plotted against the degree of ruddering, 6R, and boat-tailing, QB, where:

These can be thought of as diagonal lines running across the maps of Figure 12. While the values shown here are particular to b = 6°, and d = 13%, the trends are broadly the same for all b ¹ 0° so this data can be considered representative.

A general view of these graphs reveals that the location of the flap deflection for minimum CD and -<C_P> is different: the pair (QB, 6R) that minimizes CD is around (8°, 8°) whereas the pair the pair (QB, 6R) that minimizes -<C_P> is around (0°, 16°).

This indicates that the overall changes in CD and -<C_P>, even if related, are not governed by the same factors so other aspects beyond base suction, such as forces on flaps or circulation around the vehicle, affect the total body drag. As discussed above, for the highest reduction in drag over the operation of a vehicle, a system is required that both provides optimised drag reduction at non- zero yaw angles, and that is adaptive to changes in flow over the vehicle. To enable this, a system has been developed involving an algorithm for adapting the deflection of the control surfaces in real-time. Three types of algorithm are proposed:

1. algorithms that provide predefined flap angles from a lookup table based on real-time knowledge of the effective yaw angle; 2. algorithms that perform a real-time optimisation of flap angles, based on knowledge of the instantaneous drag; and

3. algorithms which are a combination of the two. The first type of algorithm is simplest to implement and is based on previously obtained data, such as that displayed in Figure 7. In some embodiments, such data is vehicle specific. In this case, the data may be obtained from one or more tests performed on different vehicle geometries. This data may be stored in a lookup table, in which the flap angles are chosen based upon the current estimated yaw angle, and possibly also the current speed of the vehicle.

In some embodiments, the system measures the stagnation pressure at one or more locations on the vehicle, for example using a one or more pitot tubes, to determine the yaw angle. In some cases, a single, multi-directional pitot tube (e.g. a five-hole probe) could be used. Figure 13 shows an example arrangement of sensor locations for sensors 416 at the front, along the side, and at the rear of the vehicle. For the first type of algorithm, the yaw angle determination may be made using measurements from near the front of the vehicle.

In some embodiments, the determined yaw angle is processed by a control algorithm such as that shown in figure 14. The determined yaw angle, b, may be passed through a low-pass filter fi_ to remove the effect of atmospheric turbulence and transient disturbances (e.g. from a passing vehicle). The speed of the vehicle, V, may also be input, with the filtered yaw angle, to a lookup table K. The control algorithm determines the optimal flap deflections, qi and (¼, from the lookup table. The adaptation timescale of such a system would be around 5s.

The second type of algorithm performs a real-time optimisation of the drag based upon estimates of the current gradient of the mapping between flap angles and drag. These gradients are partial derivatives dϋ/dq. An example algorithm that includes features particular to some embodiments is shown in figure 15. One such real-time optimisation method is known as extremum-seeking. Such systems consist most generally of an estimator, which estimates the instantaneous gradient, and a controller that adapts the inputs in response to this estimated gradient. The extremum-seeking system works by perturbing the input to the vehicle, in this case the flap angles, and observing the response. Since the perturbation is generally of a unique form, the response due to this perturbation is easy to isolate from the other sources of variation in the engine power. More specifically, a standard extremum-seeking system provides a sinusoidal perturbation at a known frequency, and therefore extracts measured variations at this frequency alone.

At the top of the diagram in figure 15 is a block representing the response of the vehicle between a flap angle Q and the engine power P. This is related via the nonlinear mapping N(q) and a transfer function G that captures the frequency response. Additional perturbations to the engine power are captured by w. The flap angle Q consists of a baseline value Q, a small perturbation or "dither" signal d(t), and a correction Q from the controller. The measured power P is fed into an estimation algorithm that attempts to estimate the local gradient of the mapping N(q) by detecting the response of the measured power to the known perturbation. This provides an estimate of the gradient of the mapping between flap angle and drag. i.e. the local slope of the surface shown in figure 12. One of the key benefits of using this type of algorithm is that the actual drag does not need to be known exactly, since it is only required to see how it has changed in response to small changes in the flap angle.

The engine power is affected by the vehicle acceleration or deceleration, the road surface, the gradient of the road and the mass of the vehicle, m, as well as the total aerodynamic drag. The total shaft power provided by the engine is equal to the sum of: (i) the dissipation from aerodynamic drag, (ii) the power dissipated due to rolling resistance, (iii) the rate of change of gravitational potential energy, (iv) the rate of change of kinetic energy, (v) other small losses in the power transmission e.g. gearbox. The base drag part of contribution (i) is what we wish to infer in real-time by subtracting contributions (ii-v) from the total power. Contribution (ii) is a function of the vehicle velocity and the current road surface. Contribution (iii) is a known function of the vehicle mass and the gradient of the road surface. Contribution (iv) is a known function of the rate of change of vehicle speed. Contribution (v) may be vehicle and speed dependent. The total power produced by the engine is a function of the shaft speed and mass flow rate of fuel. These contributions to engine power may be written as follows:

P = 0.5 pV³C_D + RV + mgsin(a)V + mVV + f(V)

i H iii iv v where P is the total shaft power provided by the engine, p is the density of air, V is the vehicle speed, CD is the drag coefficient, R is the rolling resistance, a is the gradient of the road surface, and V is the rate of change of vehicle speed. Most of the relevant quantities can be measured from a combination of accelerometers, GPS, and the standard parameters measured by the onboard computer (vehicle speed, engine speed, mass ow rate). For a given vehicle, the functions (ii-v) may be obtained through one or more of physical principles, detailed parametric tests, or a supervised machine learning method such as a neural network. In the latter two cases, tests may be performed on a vehicle in which the shaft power can be accurately measured. Given knowledge of each of the functions (ii-v), the real-time system would need to estimate some additional parameters such as the instantaneous vehicle mass and current coefficient of rolling resistance CR. This could be achieved via an estimation method such as the Kalman filter, that would use all the other measurements detailed above. In some embodiments, the data related to the vehicle telematics (for example power, speed, driving wheel position, brakes, etc.) can be read from the ECU 410, and other data, such as flap position or torque applied on actuators, can be collected from the flaps 418, the actuators moving them 420, or sensors embedded within the control unit 414 (GPS or accelerometers).

In other embodiments, the drag estimate is based on measurements of the pressure over the rear of the vehicle and/or forces on the flaps. The pressure at the rear of the vehicle provides a more direct measurement of the aerodynamic drag, however may require an additional array of pressure sensors.

This second type of algorithm would be more robust to changes in the vehicle without recourse to a whole new set of experiments, for example if vehicle geometry were to change from that on which the tests were performed to produce the lookup table of the first type of algorithm. However, the adaptation time of such systems is typically slower, for example around 50s. The third approach combines the advantages of each of the two preceding algorithms. This is also shown in Figure 15. In this embodiment, the lookup table provides the initial value of Q seen at the far left of the diagram. If this is not provided then Q may be chosen as the optimal flap angle under average conditions. The general principle is to select an initial set of flap angles based on the lookup table, and to additionally adapt these angles in response to real-time estimates of the drag. This would achieve both a more rapid adaptation time (5s) but also be robust to differences between the real vehicle and that on which experiments were performed. In some embodiments, the system may further update the lookup table based upon the optimal configurations found in practice. For example, methods such as Gaussian process regression and Bayesian inference may be used to generate the updated mapping.

The drag reduction approach presented herein has the advantage of delivering drag reduction with minimal interference on the body geometry of the bluff body. This distinctive feature of minimal geometric modification is beneficial for a number of reasons. Specifically, the drag reduction approach set out in this application is beneficial over other technologies, since it can be installed on a vehicle or other bluff body within restrictive legal constraints on geometry. The present disclosure aims to modify the aerodynamic forces and moments over a bluff body, such as a road vehicle, by positioning control surfaces located towards the vehicle rear end of the bluff body about a pivot point. For example, wherein the object is a vehicle, the rear end may be the end of the vehicle facing away from the direction of travel.

The present disclosure relates to reducing the aerodynamic drag. Nevertheless, the principles set forth in this disclosure may allow for control over the surfaces so that drag can be increased (to improve braking effectiveness) or the lateral forces can be varied (to help in turns, if necessary). Drag is reduced specifically by interacting with environmental flow asymmetries (such as cross-wind or any asymmetric operational configuration). The described methods and control algorithms may generally be implemented by a computer program. The computer program may be in the form of computer- executable instructions or code arranged to instruct or cause a computer or processor to perform one or more functions of the described methods. The computer program may be provided to an apparatus, such as a computer, on a computer readable medium or computer program product. The computer readable medium or computer program product may comprise non-transitory media such as semiconductor or solid-state memory, magnetic tape, a removable computer memory stick or diskette, a random-access memory (RAM), a read- only memory (ROM), a rigid magnetic disc, and an optical disk, such as a CD- ROM, CD-R/W, DVD or Blu-ray. Another implementation would be the use of Field-Programmable Gate Array (FPGA) technology wherein a commercially available integrated circuit can be programmed for this specific application. The computer readable medium or computer program product may comprise a transmission signal. An apparatus or device such as a computer may be configured to perform one or more functions of the described methods.

Other variations and modifications will be apparent to the skilled person. Such variations and modifications may involve equivalent and other features which are already known and which may be used instead of, or in addition to, features described herein. Features that are described in the context of separate embodiments may be provided in combination in a single embodiment. Conversely, features which are described in the context of a single embodiment may also be provided separately or in any suitable sub-combination. It should be noted that the term "comprising" does not exclude other elements or steps, the term "a" or "an" does not exclude a plurality, a single feature may fulfil the functions of several features recited in the claims and reference signs in the claims shall not be construed as limiting the scope of the claims. It should also be noted that the Figures are not necessarily to scale; emphasis instead generally being placed upon illustrating the principles of the present disclosure.

Claims

1. A drag reduction apparatus for reducing the aerodynamic drag on a bluff body with a blunt trailing edge caused by fluid flow characteristics at the wake of the bluff body, the drag reduction apparatus comprising:

one or more control elements configured to be coupled to the bluff body and to be positioned asymmetrically relative to a direction of travel of the bluff body in response to a determined parameter of the bluff body to control fluid flow at the wake of the bluff body and reduce the drag experienced by the body.

2. The drag reduction apparatus according to claim 1 , further comprising one or more actuators configured to cause the one or more control surfaces to move with respect to the bluff body.

3. The drag reduction apparatus according to claim 1 or 2, further comprising a processor configured to execute a control algorithm so as to generate at least one control signal for controlling the position of the one or more control surfaces.

4. The drag reduction apparatus according to claim 3, wherein the position of the one or more control surfaces is a rotational position.

5. The drag reduction apparatus according to claim 3 or 4, wherein the processor is configured generate the control signal based on the parameter of the bluff body.

6. The drag reduction apparatus according to claim 5, wherein the drag reduction apparatus further comprises at least one sensor configured to determine an environmental parameter, and wherein the processor is configured to determine the parameter of the bluff body based upon an output of the at least one sensor.

7. The drag reduction apparatus according to claim 6, wherein the environmental parameter is the stagnation pressure at one or more locations towards the front of the bluff body.

8. The drag reduction apparatus according to claim 7, wherein the at least one sensor is at least one pitot tube.

9. The drag reduction apparatus according to claim 7 or 8, wherein the processor is configured to apply a low pass filter to the output of the sensor.

10. The drag reduction apparatus according to any of claims 7 to 9, wherein the processor is configured to determine the effective yaw angle of the bluff body based upon the determined stagnation pressure.

11. The drag reduction apparatus according to claim 10, wherein the processor is configured to generate the control signal based on the determined effective yaw angle and a lookup table of effective yaw angles and control surface positions.

12. The drag reduction apparatus according to claim 5, wherein the parameter off the bluff body is the drag currently experienced by the bluff body.

13. The drag reduction apparatus according to claim 12, wherein the processor is configured to determine the drag based on the current pressure at the rear of the body.

14. The drag reduction apparatus according to claim 12 or 13, wherein the processor is configured to determine the drag based on a force currently experienced by the one or more control surfaces.

15. The drag reduction apparatus according to any of claims 12 to 14, wherein the bluff body is driven by an engine and the processor is configured to determine the drag based on the current engine power.

16. A drag reduction apparatus according to any preceding claim, wherein the bluff body is a road vehicle.

17. The drag reduction apparatus according to claim 16, wherein the road vehicle is a heavy goods vehicle“HGV” or a light goods vehicle“LGV”.

18. A vehicle comprising the drag reduction apparatus of any preceding claim.

19. A vehicle according to claim 18, comprising a plurality of pressure sensors positioned on the vehicle so as to measure pressure in the wake of the vehicle.

20. A vehicle according to claim 18 or 19, comprising a plurality of control surfaces, each coupled to the vehicle at a rear of the vehicle so as to control the fluid flow at the wake of the vehicle.

21. A drag reduction method for reducing the aerodynamic drag on a bluff body with a blunt trailing edge caused by fluid flow characteristics at the wake of the bluff body, the drag reduction method comprising:

coupling one or more control elements to the bluff body; determ ining a parameter of the bluff body; and

positioning the one or more control elements asymmetrically relative to a direction of travel of the bluff body in response to the sensed parameter to control fluid flow at the wake of the bluff body and reduce the drag experienced by the body.

22. The drag reduction method according to claim 21 , further comprising moving the one more control surfaces by moving one or more actuators configured to cause the one more or more control surfaces to move to control fluid flow.

23. The drag reduction method according to claim 21 or 22, further comprising executing a control algorithm using a processor to generate at least one control signal for controlling the one or more actuators.

24. The drag reduction method according to claim 23, further comprising controlling a rotational position of the one or more control elements.

25. The drag reduction method according to claim 23 or 24, further comprising generating the control signal based on the parameter of the bluff body.

26. The drag reduction method according to claim 25, further comprising determining the parameter of the bluff body by detecting an environmental parameter using at least one sensor.

27. The drag reduction method according to claim 26, wherein the environmental parameter is the stagnation pressure at one or more locations towards the front of the bluff body.

28. The drag reduction method according to claim 27, wherein the at least one sensor is at least one pitot tube.

29. The drag reduction method according to claim 27 or 28, further comprising applying a low pass filter to the output of the sensor.

30. The drag reduction method according to any of claims 27 to 29, further comprising determining the effective yaw angle of the bluff body based upon the determined stagnation pressure.

31. The drag reduction method according to claim 30, further comprising generating the control signal based on the determined effective yaw angle and a lookup table of effective yaw angles and control surface positions.

32. The drag reduction method according to claim 26, wherein the parameter off the bluff body is the drag currently experienced by the bluff body.

33. The drag reduction method according to claim 32, further comprising determining the drag based on the current pressure at the rear of the body.

34. The drag reduction method according to claim 32 or 33, further comprising determining the drag based on a force currently experienced by the one or more control surfaces.

35. The drag reduction method according to any of claims 32 to 34, further comprising driving the bluff body with an engine and determining the drag based on the current engine power.

36. A drag reduction method according to any of claims 21 to 35, wherein the bluff body is a road vehicle.

37. The drag reduction method according to claim 36, wherein the road vehicle is a heavy goods vehicle“HGV” or a light goods vehicle“LGV”.

38. A computer-readable medium comprising computer-readable instructions which, when executed by a processor, cause the processor to perform the method of any of claims 20 to 37.