Traditionally, active magnetic bearing (AMB) systems are designed as an integral component of machines having generally complex dynamic characteristics. An AMB supported rotor has been tested over a speed range that included system natural vibration frequencies. A linear stiffness and viscous damping AMB characteristic with constant coefficients was identified which was independent of the overall system characteristics and which can thus provide simple and transferable data for a machine designer.

The use of computational fluid dynamics (CFD) software for modelling the flow of electro-structured fluids is introduced. A non-Newtonian fluids package written specifically to model Bingham plastics is validated for several flow rates between stationary parallel plates for varying yield stresses, plate separations and lengths. The computing procedure is rationalised in terms of grid fitting of the 'plug' edge. The programme is modified to include an analytical expression which relates delectro-rheological fluid parameters. This approach is then used to predict valve flow rates from small sample, Couette viscometer produced data: its output compares with experimental results.

The performance of electric vehicles may be improved significantly by incorporating a flywheel peak power buffer into the power-train. An annular carbon fibre composite flywheel rim offers a high specific energy capability, although a high rotational speed is necessary for a high stored energy, whilst operation in a vacuum is conducive to a low aerodynamic loss, and necessitates the use of magnetic bearings. The electromagnetic design of such a flywheel system is described, and a dynamic analysis that predicts the free vibration modes and frequencies, is presented. Finally, the magnetic bearing control system is described, and it is shown that the vibration damping capability of the bearings is such that natural vibration frequencies within the flywheel operating speed range can be accommodated, and performance objectives commensurate with operation in a vehicle drive-train can be achieved.

Semi-active damping devices offer improved performance over passive devices, without the power requirements or instability problems of fully active devices. Smart fluids (electrorheological and magnetorheological) are well suited to use in semi-active dampers—their flow properties can be rapidly altered, with a low-power requirement. However, the force/velocity response is highly non-linear, and this is without doubt hindering the development of effective control strategies. In this paper, the authors develop a new model of an electrorheological damper. The key advantage of this model is that its algebraic form is suitable for use in control system design, whilst it is able to predict and explain observed behaviour. The model consists of a spring, mass, and damper connected in series. The spring stiffness term is based upon the fluid bulk modulus, and the mass is determined from the fluid density. The damping characteristic utilizes a modified non-dimensional Bingham Plastic function. The model predictions are compared with experimental results at a range of operating frequencies. Excellent agreement was achieved by updating the stiffness and viscosity parameters using experimental data.

It is now well known that smart fluids [electrorheological (ER) and magnetorheological (MR)] can form the basis of controllable vibration damping devices. With both types of fluid, however, the force/velocity characteristic of the resulting damper is significantly non-linear, possessing the general form associated with a Bingham plastic. In a previous paper the authors showed that by using a linear feedback control strategy it is possible to produce the equivalent of a viscous damper with a continuously variable damping coefficient. In the present paper the authors illustrate an extension of the technique, by showing how the shape of the force/velocity characteristic can be controlled through feedback control. This is achieved by using a polynomial function to generate a set point based upon the damper velocity. The response is investigated for polynomial functions of zero, 1st and 2nd order. It is shown how the damper can accurately track higher order polynomial shaping functions, while the zero-order function is particularly useful in illustrating the dynamics of the closed-loop system.

It is widely acknowledged that the inherent non-linearity of smart fluid dampers is inhibiting the development of effective control regimes, and mass-production devices. In an earlier publication, an innovative solution to this problem was presented -- using a simple feedback control strategy to linearize the response. The study used a quasi-steady model of a long-stroke Electrorheological damper, and showed how proportional feedback control could linearize the simulated response. However, this initial research did not consider the dynamics of the damper's behavior, and so the development of a more advanced model has been necessary. In this article, the authors present an extension to this earlier study, using a model of the damper's response that is capable of accurately predicting the dynamic response of the damper. To introduce the topic, the electrorheological long-stroke damper test rig is described, and an overview of the earlier study is given. The advanced model is then derived, and its predictions are compared to experimental data from the test rig. This model is then incorporated into the feedback control simulations, and it is shown how the control strategy is still able to linearize the response in simulations.

It is now well known that smart fluids (electrorheological (ER) and magnetorheological) can form the basis of controllable vibration damping devices. With both types of fluid, however, the force/velocity characteristic of the resulting damper is significantly nonlinear, possessing the general form associated with a Bingham plastic. In a previous paper the authors suggested that by using a linear feedback control strategy it should be possible to produce the equivalent of a viscous damper with a continuously variable damping coefficient. In the present paper the authors describe a comprehensive investigation into the implementation of this linearization strategy on an industrial scale ER long-stroke vibration damper. Using mechanical excitation frequencies up to 5 Hz it is shown that linear behaviour can be obtained between well defined limits and that the slope of the linearized force/velocity characteristic can be specified through the choice of a controller gain term.

Electrically-structured (ES) fluids offer a potentially elegant means of introducing greater flexibility into a range of industrial machines and structures, and are especially suited to controllable vibration damping. The authors have developed a technique for characterising ES fluids, leading to a mathematical model and practical device design procedure, the principles of which are verified by testing with an industrial scale, long-stroke electro-rheological (ER) damper, suited to the needs of a rail vehicle lateral suspension. In this paper, the model is applied to predict the effect on the test damper steady state performance, and performance envelope: of the principle variables of operating temperature, mechanical displacement amplitude and frequency, and control field excitation, which influence the controllability and control requirements of the damper, and to show that the test damper should achieve the desired range of control.

A method for the rapid solution of flows of electrically structured fluids of the particle/liquid type is described. The technique is indicated for heavily loaded fluids and for flows in complex controller geometries. Its basis lies in the modification of a commercial computer routine which can deal with Bingham plastics, by the inclusion of a sub-routine which represents the ERF/MRS constitutive equations for a wide range of excitation. The entire system of working is validated against known and well proven analytical and experimental results.

Traditionally, procedures of modelling the performance of electro-rheological (ER) devices have been severely hampered by the absence of robust methods for characterising the behaviour of the ER fluids themselves. Recent publications by the authors have described a non-dimensional characterisation procedure. Not only does this new approach drastically reduce the number of variables but through a generalised analysis enables tire steady flow characteristics of different types and sizes of devices to be related. In this paper, the authors describe the extension of the characterisation technique to consider the behaviour of a controllable vibration damper which uses ER fluid as the working medium. A critical comparison is made between the model predictions and the observed behaviour of a test facility specially constructed to examine the response of an ER damper to harmonic displacement excitation. Emphasis is placed upon the influence of fluid inertia and compressibility on dynamic response and upon phenomena which are observed at the reversal of fluid motion; suggestions are made for further work.

The development of controllable suspension dampers for ground vehicles is the subject of much current research. In this paper the authors describe aspects of a design methodology for controllable dampers which use electro-rheological (ER) fluid as the working medium. This methodology is based upon a non-dimensional characterisation of ER fluid data which allows measurements obtained from small-scale tests to be used to predict the behaviour of industrial-scale vibration dampers. The ER damper is represented via a Bingham plastic constitutive relationship, augmented by terms to account for fluid inertia and compressibility. An industrial-scale test facility is described and the first available set of experimental results are presented. A comparison is made between model predictions and observed behaviour.

An effective design methodology for ER devices requires reliable data to characterize the behavior of the ER fluid as distinct from the test device. In this paper the authors describe a technique they have used for generating ER fluid data so that it can be used to predict the performance of different types and sizes of device operating with the test fluid. The application of the technique is then illustrated using a problem of current industrial interest: namely the dynamic modeling of a controllable vibration damper for vehicle suspension applications.

In this paper the authors consider the problem of optimising the design of a magneto-rheological (MR) damper for controlling the lateral vibrations of a modern rail vehicle. It is shown how, using generalised test data, the primary control variable (i.e. the applied magnetic field) can be related to the desired force-velocity envelope of the MR damper. Optimisation of the damper design is then considered, taking account of the damper dimensions and available space envelope, the dynamics (due to compressibility and inertia) of the MR fluid and the configuration of the magnetic circuit. Using data from a commercially available MR fluid, it is shown how a controllable damper can be designed to improve upon the specification for a conventional lateral railcar damper. With zero applied magnetic field the MR damper is shown to be capable of providing a fail-safe minimum value of damping. Increasing the magnetic field provides a maximum equivalent damping level in excess of that available from a conventional hydraulic damper. Over the frequency range from 1 to 10 Hz it is suggested that dynamic effects would not cause significant degradation of performance. Furthermore, for the MR fluid considered, the electrical power consumption would be unlikely to exceed 100 watts.

In previous papers a technique for comparing the performance of a particular electrorheological (ER) fluid in different situations was derived from steady, laminar, shear and flow mode experimental data and static cell tests. By plotting non-dimensional Hedström Number (He) and Friction Coefficient (Cf) against Reynolds Number (Re) via use of the Bingham plastic constitutive equation, one continuous and overlapping relationship was shown for constant field excitation over a range of operating conditions and geometries. In the present work the same treatment is extended to a different fluid, of higher solid volume fraction and different constituents, which was tested on different instruments at a different uniform temperature, so as to attest the generality of the technique. In doing this the methodology of converting low shear rate Couette viscometer data to practical flow valve performance predictions is illustrated.

In this paper, the authors describe the development of a mathematical model of a controllable vibration damper intended for eventual application to ground-vehicle suspension systems. The damper under investigation employs electro-rheological (ER) fluid as the working medium which enables a continuously variable damping force to be provided in response to an electrical control signal. There are some difficulties inherent in characterizing the ER damper's behaviour which the present study attempts to overcome.The paper begins by describing a novel form of non-dimensionalization which drastically reduces the number of variables required to characterize the quasi-steady behaviour of the ER fluid. The construction of the ER damper is described and, on the basis of physical reasoning, it is shown how a dynamic model can be derived by taking account of ER fluid inertia and compressibility. A recently developed iterative scheme is introduced in order to solve the resulting non-linear equations of motion. The paper concludes with a case study involving the application of the ER damper to controlling the lateral vibrations of a rail vehicle.

We examine the experimental pressure response for a set of electrorheological valves operating at several steady flow rates to step function voltage inputs and analyse the factors that may contribute to the complex forms observed. Using a numerical algorithm incorporating a viscoelastic model for the fluid the response of the valve flow profile to changes in rheology is found to be relatively fast compared to the experimental pressure response and thought not to feature as a significant contribution to the final shape. The dynamic interaction of the valve and its associated system of pump and connecting pipes is investigated using a multi-element lumped parameter model and reveals a frequency spectrum containing resonances comparable to the experimental results. A simple lumped parameter model is less successful. Using the multi-element model for the system an upper bound can be placed on the basic response time of the electro-stress itself under realistic engineering conditions.

An earlier model describing the electrical characteristics of an electrorheological (ER) valve is refined to include the effects of coupling to the pressure response. Expressions including additional linear terms in the pressure and time differential of the pressure are considered and it is found that a single term in the time derivative with a small time delay is quite a sufficient description of the data on the basis of goodness of fit.

Experimental data from tests on an ER valve pertinent to the development of a controller for high speed machine duty has been analysed to show that three common, underlying modes of response are present. This is demonstrated for a range of industrial scale flow velocities and electrode dimensions, in the time and frequency domains. The dependence of steady-state pressure on the electric field is also discussed.

Yield shear stresses and shear rates at a valve wall are derived from experimental results on a series of concentric cylinder valves. The fluid constitutive equation used for this purpose is that of a Bingham Plastic. Valve plates (which are such that the radial gap is small compared to its mean pitch) are taken to be parallel so far as the derivation of the flow ν's pressure ν's geometry model is concerned. A range of electrode separations from 0.5 to 1.0mm are used with flow velocities being limited to the region where the viscous pressure drop component is below that caused by the electro stress. Results show that (away from the region of low shear rates and high voltages) the wall stresses for equivalent conditions are comparable for different valves, for a range of applied field strengths and mean flow velocities. Thus, provided the hysteretic region is avoided the fluid can be treated as a Bingham continuum with some stated reservations. However, this is only applied with precision for the truly corresponding situations defined in the paper.