lt is the author's contention that a brake control system incorporating the following two features is suitable for high-performance, robust and adaptive anti-skid control. 1. Switching on wheel acceleration with hysteresis, 2. Fixed brake pressure rate control. Furthermore, the author contends that such a system can be developed in the context of large commercial aircraft development and with regard to; Minimal a-priori knowledge of system dynamics, I No requirement for on-line system identification. This thesis documents a programme of research conducted to examine this hypothesis in the context of large commercial aircraft, and in conjunction with Airbus UK at Filton in Bristol. The physical basis for this method of control is examined from first principles, explored in detail through a simulation based experimental process and is subsequently confirmed as a viable means of control. This is further developed into a prototype anti-skid braking control algorithm which in turn is incorporated into a nominal brake control system model. A detailed analysis of the systems performance is then conducted and reported upon. Finally, conclusions are drawn with respect to the aims and objectives documented herein and recommendations for further research are made. The novel aspects of this hypothesis and subsequent algorithm development are as follows, and are to be considered with respect to the current state-of-the-art in the field: 1. The use of a fixed brake pressure rate control to isolate the friction dependent aspects of the system dynamics, 2. The use of a simple switching element which incorporates hysteresis such that when combined with l, allows robust -slip gradient control* to be effected, and 3. The use of brake system response characteristics to drive an optimal wheel acceleration demand signal. The author believes that these three aspects of the anti-skid braking algorithm documented within this thesis combine to provide an effective and pragmatic solution to the problem, and represent a step forward in the state of- the-art as is befitting a research programme of EngD character. 'The gradient of the -slip (friction coefficient vs. wheel slip ratio) curve is controlled.

Magnetorheological (MR) fluids enable the rapid and continuous alteration of flow resistance via the application of a magnetic field. This unique characteristic can be utilised to build semi-active dampers for a wide variety of vibration control systems, including structural, automotive, and aeronautical applications. As an example, MR fluids could enhance the performance of aircraft landing gear, which are subject to widely varied and unpredictable impact conditions with conflicting damping requirements. In this thesis, a numerical sizing methodology is developed that enables the impact performance of MR landing gears to be optimised. Using real data provided by landing gear manufacturers, the sizing methodology is applied to both lightweight aircraft, and large-scale commercial jets in order to demonstrate scalability. For both aircraft types, results indicate that the peak force and the severity of fatigue loading can be enhanced over a wide range of impact conditions. However, it is shown that MR landing gears can be heavier than passive systems. To validate the numerical approach, a prototype MR landing gear shock strut is designed, fabricated, and tested. Good correlation between the model and experiment is demonstrated, particularly for low velocity excitations. MR dampers exhibit highly non-linear force-velocity behaviour. For landing gear impacts, it transpires that this behaviour can be used to an advantage, where it is shown that an acceptable performance can be obtained using open-loop control i.e. with a constant magnetic field. However, this non-linear behaviour is highly undesirable for other scenarios (e.g. an aircraft taxiing), and as a consequence, the choice of an effecti\'e control strategy remains an unresolved problem. A further aim of this thesis is therefore to develop effective control techniques for broadband excited MR vibration systems. Through an extensive series of numerical and experimental investigations, case studics representative of the general single-degree-of-freedom and two-degree-of-freedom vibration isolation problem are presented. In the experiments, the hardware-in-the-Ioopsimulation method is adopted, which provides an excellent means to bridge the gap between theory and practice when the behaviour of a specific component is complex. Here, the MR damper is physically tested, whilst the remainder of the structure is simulated in real-time. The results demonstrate that the chosen control strategy can provide significant performance benefits when compared to more commonly used strategies and equivalent passive systems. Furthermore, the control strategy is shown to be insensitive to factors such as the type of input excitation.

This thesis presents a bifurcation and numerical continuation study into the occurrence of shimmy instability in an aircraft main landing gear (MLG) of single side-stay, dualwheel design. The dynamics are expressed in terms of three rotational degrees of freedom (Dofs) aligned with the side-stay plane, and a fourth translational DoF representing compression of the main strut. These DoFs are modelled by oscillators that are coupled directly through the geometric configuration of the system, as well as through the tyre/ground interface. Using this representation of the MLG system we focus in this thesis on the nonlinear effects of geometric orientation and mechanical freeplay employing bifurcation analysis techniques to highlight their effects on the MLG stability. First, a changing side-stay orientation angle is investigated. Shimmy is studied by means of a two-parameter bifurcation analysis in terms of the landing gear forward velocity and loading force. For this a three-DoF model is used that does not include axial compression. This formulation along with suitable parameters allows for comparison with the existing literature, and an agreement is demonstrated with previous results for a zero side-stay angle. Subsequent variation of this angle is explored and a consistent picture presented, capturing the (transition of the two-parameter bifurcation diagram us a function of this angle. This shows a considerable increase in the complexity of the dynamics for intermediate side-stay angles. The appearance of an additional shimmy mode is observed and a region of tri-stability found where three distinct shimmy types coexist. For the study of freeplay the MLG model is extended to include axial compression; this axial DoF is required to accurately represent freeplay, introduced to the torque links of the system. Parameter values are chosen here to represent a typical mid-range civil aircraft MLG. The addition of freeplay is shown to result in shimmy oscillations that occur within the MLC operating envelope; their properties depend on both the size and 'shape' of freeplay. Freeplay and geometric coupling arc also considered together via consideration of a non-zero side-stay angle. Here, additional dynamic complexity is introduced in the presence of freeplay and, again , this coincidence with the appearance of a new shimmy mode. Further complex phenomena also appear, including multiple- frequency and chaotic-type oscillations, as well as complex transients. The non-zero geometry produce asymmetry and this results in a great sensitivity of the small-amplitude MLG behaviour to the exact shape of freeplay. Therefore, geometric orientation and freeplay a.re found to have significant effects and, when combined, they work together to produce additional complex phenomena, not otherwise observed when considered in isolation.

Modem aircraft use Carbon-Carbon Composite (CCC) materials for the heat-pack elements of the brake (Tarter, 1991). The elevated wear-rate mechanism of CCC brakes at low operating temperatures (<100°C) is well understood (Abbott, 1995). Oxidation of these friction materials increases between temperatures of 400°C - 600°C (Cullinan et al, 1989), which also increases the level of heat-pack wear. The current method for monitoring the life of CCC brakes is by visual inspection of the wear pin indicator. The limitation of this method is that there is no means of determining the causes of elevated heat-pack wear. This executive summary presents two innovative solutions to this wear-rate problem. The first solution monitors and records the amount of low-energy braking applications that occur on a conventional hydraulic braking system by processing the representative temperature and pressure information of the hydraulic brake fluid.

No description available.

A research programme has been conducted, the objective of which has been the determination of unsteady force coefficients for a range of parachute canopy models. These coefficients are required for prediction of the aerodynamic stability of full scale parachutes under conditions of unsteady motion during descent. The method of obtaining these coefficients required the collection of force and acceleration data for parachute canopy models which were tested in unsteady conditions. This was achieved by imposing oscillatory motion on individual canopies during towing tests, which were conducted under water in a ship testing tank. Two modes of unsteady motion were imposed on a canopy under test; one in which it was oscillated along its axis, and one in which it was oscillated laterally. A mathematical model describing such modes of motion consists of a general equation for the unsteady force developed on a bluff body. In this model the force F(t) is expressed using two components; a velocity dependent force component, and an acceleration dependent force component. Each component of the aerodynamic force contains an unknown parameter denoted by the terms ‘a’ and ‘b’ in the equation, which is shown below; F( t ) = a( t ) • V²( t ) + b( t ) • V( t ). An identification technique is used to determine the mean values per cycle of each parameter by substitution of the data obtained from these tests as functional variables in the mathematical model. Mean values of the velocity dependent force and stability coefficients; CT and ∂CN/∂α, and the added mass coefficients k11 and k33 are then obtained from these parameters. The results of this programme indicate a strong dependence in oscillatory motion of the mean value per cycle for the axial added mass coefficient k11 on the unsteady force parameter called the Keulegan-Carpenter number KC; KC = Û • T/DO. Where; Û = the velocity amplitude of the oscillation, T = the period of an oscillation, and DO = a typical canopy dimension. The velocity dependent axial force coefficient CT exhibits a similar, although not as substantial dependency. Good agreement has been obtained between steady-state test results from this programme and results from other independent work. The effects of values obtained in this investigation are considered in the linearised dynamic stability model developed by Doherr and Saliaris (1), and their influence on the descent characteristics of full-scale parachutes is assesed.