Search

The analysis of externally pressurized, plane, oil-film pads has been mainly restricted, in the current literature, to simple pad shapes without tilt or relative sliding action or squeeze effects. As tilt can arise owing to manufacturing tolerances, or by dynamic inertial forces causing elastic deformation, it is important to assess the combined effects of tilt and sliding action in order to predict such factors as the stability of the system or the onset of cavitation. This paper presents general numerical solutions from the Atlas digital computer of the Reynolds equation applied mainly to circular pressurized pads. Results for a square pad and for an infinitely wide pad are also given. The load and location of the centre of pressure, the impedance of the pad and thus the flow rate are calculated in dimensionless form for any combination of tilt, sliding velocity, and squeeze rate. It is argued that these calculated terms are required before any dynamical problem, involving loaded hydrostatic bearings, can be attempted and before refinements of film lubrication theory can be incorporated. This paper illustrates the application of the results to dynamical problems.

This paper contains the results, plotted non-dimensionally, of extensive tests carried out on a 1–625-in diameter grinding machine spindle lubricated hydrostatically. The horizontal spindle is directly coupled to a variable-speed motor at one end and loaded vertically at the other. The spindle is not constrained to rotate at arbitrary eccentricity values, but is able to seek an equilibrium position depending upon oil flow and rotational speed. The attitude, oil film thickness, and eccentricity are measured by mutually, perpendicular capacitance probes arranged in the bearing housing. The results indicate considerably decreasing load-carrying capacity as the rotational speed increases, because the hydrodynamic effect tends to centralize the shaft in the bearing housing, thus nullifying the hydrostatic effect. Paradoxically the same effect causes an increase in oil film stiffness as it is generally defined. A summary of earlier work is included, notably that of Raimondi and Boyd who considered bearings of fixed eccentricity, and the paper also presents a theoretical analysis based on the pressure- and velocity-induced circumferential flow theories.

The objects of this paper are to demonstrate the use of the method described in Paper 6 and to give an indication of the effect of design parameters on the performance of a dynamically loaded bearing. An investigation into the main and big-end bearings of a diesel engine of modern design is described. The performance of a bearing is gauged by obtaining: (a) the minimum oil film thickness, (6) the maximum pressure developed within the oil, (c) the maximum load/projected area, and ( d) the bulk oil temperature rise. The influence of the engine design upon the big-end bearing performance for the particular engine under consideration is found by investigating the effects of the following parameters: ( a) bearing length-diameter ratio, (6) removal of centre-circumferential groove, ( c) inlet temperature, ( d) bearing clearance, and ( e) loading diagram. This last parameter is varied by changing in turn the piston mass, the connecting-rod mass, and the position of the connecting-rod centre of gravity. Finally, the effect upon the main bearing performance of various balance weight configurations is demonstrated. The usual assumption is made to obtain the main bearing loads: that loads imposed by a crank throw are equally distributed between neighbouring main bearings. General conclusions are made regarding the relative influence of certain bearing-design parameters upon the bearing performance.

This paper describes two different computer programs for the calculation of rolling bearing applications in an automotive driveline. The first program gives the effect of bearing preload or internal clearance on bearing deflections and nominal lives in bearing applications with taper roller bearings. As an example, the paper deals with a front hub bearing arrangement with two taper roller bearings used in a front wheel driven car, and also with a pinion of a final drive unit and a gearbox mainshaft incorporating two taper and one cylindrical roller bearing. The second program is used to determine the influence of misalignment and shaft deflection on the load and life of a deep groove ball bearing. To illustrate this effect the layshaft of an automotive gearbox is studied.

The design of a rig to determine the load and attitude relationships of a hydrodynamically lubricated bearing under steady and dynamic loading is described. The bearing is nominally of 3 in diameter with a diametral clearance of 0·0075 in. Two bearing lands of length-diameter ratio of 0·282 are fed by a central circumferential oil groove. It has been designed on these model conditions to minimize bearing distortion and temperature gradients in order to provide a basis for comparison with the results of the computer program reported in Paper 6. Two perpendicular load components are applied mechanically through spring levers displaced by cams. The load components are measured by strain gauges and the displacement of the bearing is detected by four inductive pick-ups. The arrangement of these permits the relative linear and angular displacements of the journal and bearing axes to be determined. Inlet and outlet oil temperatures are measured for viscosity determination. Both steady and dynamic loads are applied by the loading system. Load diagrams of approximately circular and elliptical form may be obtained directly by adjusting the magnitude and phase of the eccentrics serving as cams, while more complex load patterns may be obtained by special cams. The polar load diagrams recorded are used as data for the computer program of Paper 6. Theoretical and practical results are then compared.

At partial and overload conditions, radial decentralizing forces act upon the rotor of a centrifugal pump with guide vanes if the impeller is out of centre. The magnitude of these forces depends on load conditions, and the forces increase with growing eccentricity. At very small flow, these forces become non-stationary. They rotate at a considerably lower frequency than the velocity frequency and may lead to rotor vibrations. The paper discusses the effects of stationary and non-stationary radial forces, and resulting shaft deflections and vibrations, from measurements on two experimental test rigs. The paper also presents the results of research on the pressure distribution of guide vanes, carried out in air, and gives observations of flow patterns in a shallow water test rig.

This paper deals with ball and cylindrical roller bearings in drivelines where, year by year, requirements demand that smaller bearings meet conditions of steadily increasing severity. Clutch-operating bearings and those for manual gearboxes and overdrives are reviewed. The different requirements for bearings in front-wheel drives and automatic gearboxes are discussed, followed by a consideration of the requirements of propeller shafts, half-shafts and hubs.

Thin cylindrical shells subjected to high bending loads normally buckle by forming a number of diamond- shaped panels on part of their compressed surfaces. The values of the buckling loads show considerable scatter but are in general substantially less than those predicted by simple linear theory. This paper describes an attempt made a few years ago to improve the buckling strength and to reduce the scatter in the buckling loads of such shells by means of internal pressure. It was expected and confirmed that internal pressure would increase the buckling loads, but little improvement was obtained in the amount of scatter. Considerable modifications of the buckling deformations were observed; at sufficiently high pressures the diamond pattern disappeared entirely and was replaced by a single ridge of deformation extending circumferentially around the compressed side of the shell. An account is given of the experimental investigation together with some excerpts from the theoretical work. Most of the shells were formed from thin steel sheets bent into the shape of a cylinder then soldered down a generator. They were filled with water, plugged at their ends and put under pressure by means of a hand-pump. Observations made during bending included load, pressure, strain, deflections and mode of buckling. These are discussed and compared with the results of other investigations.

In support of the vast amount of stress analysis work on components, considerable effort is required to provide suitable materials behaviour models. Simple laboratory tests on uniaxially loaded specimens under constant temperature condition provide the bulk of data, but methods are required to translate these data to multiaxial conditions with non-uniform cyclic loading and cyclic temperature. The material behaviour laws, therefore, are aimed at meeting two requirements: first, to predict the deformamation response in terms of the imposed loading conditions and the previous history of the material; and second, to analyse the stress or strain cycles to predict failure. This paper reviews the steps being taken at Berkeley Nuclear Laboratories, C.E.G.B., and elsewhere to achieve these two objectives.

The dynamics of existing railway vehicles on straight track is dominated by motions in the lateral plane commonly referred to as hunting. Hunting is a self-excited oscillation caused by the combined action of the conicity of the wheels and the creep forces acting in the contact plane between wheel and rail. In this paper existing and recently developed theories of the lateral motion of railway vehicles are discussed in relation to experimental work on both models and full-scale vehicles. It is shown that a realistic theory taking into account flexibility between the wheelsets and the frame in the longitudinal, lateral, and vertical directions, and the influence of wheel and rail profiles, yield values for the critical speeds which are consistent with experimental results. The influence of various parameters on stability is discussed and it is explained how railway vehicles can be designed for stable running at high speeds. Examples of suitable choices of suspension and other parameters will be given and experiments carried out in order to verify these concepts will be described.

This paper describes a theoretical and experimental study of the non-steady flow performance of a centrifugal compressor. The experimental work was designed to study the effect of pulse frequency on the compressor performance, for a given delivery configuration, using a rotary valve pulse generator. The experimental rig was designed so that it was possible to study the reverse flow and the pressure pulsations in the suction side of the compressor. The objective of the investigation was to discover how the compressor performance deteriorated with pulse frequency, and also to determine the frequency at which reverse flow through the compressor first occurred. The objective of the theoretical work was to predict the onset of reverse flow through the compressor and the mean and transient delivery pressure ratio using the conventional stationary pipe non-steady solutions by the method of characteristics. The compressor unit was replaced by a boundary condition within the pipe system equal to the experimentally known steady flow characteristics of the compressor. The physical size of the compressor was replaced by an equivalent pipe length; this technique is described. The theoretical results are compared with an extensive series of experimental results. This work is a direct extension of that given in reference (3).

The experimental work reported in this paper has shown that it is possible to proceed into and beyond the half-speed whirl region and operate at much higher speeds without vibration, using plain cylindrical journal bearings. Journals of 1-in diameter, clearance ratios greater than 1 in 1000, and bushes with length-diameter ratio of 1 were used. The bushes were mounted flexibly in rubber O-rings.

This paper discusses the dynamic behaviour of two designs of railway vehicle, taking into account the non-linear features of their suspensions and of the forces arising between rail and wheel. The parameters required for the theoretical analyses are obtained from measurements of the vehicles and from detailed studies of rail and tyre profiles. The vehicle behaviour is then predicted and these predictions are checked against running tests made on a special test track. The main features discussed are the lateral motion on plain track (hunting) and the response to an isolated vertical irregularity (a ‘dipped’ rail-joint). The effect of various parameters on this behaviour is described and some comments are made upon the problems of derailment.

This paper is concerned with the machinery for the launching of aircraft from and the recovery of aircraft on to the flight decks of aircraft carriers. The machinery to do this is hydraulic—except for the working fluid of slotted cylinder catapults, which is steam—and is required to control the application of large amounts of energy, in a short period, precisely in accordance with a set programme, with complete consistency and absolute reliability. Hydraulic machinery has successfully met all these requirements for fifty years, being developed and revised to keep pace with the increases in the performance of carrier-borne aircraft. The major developments—catapults for launching and arresting gear for recovery—are described, together with some items of ancillary equipment that demonstrate the versatility of hydraulic machinery. Simple principles have always been used, but the designs have always made use of the most advanced techniques of hydraulic engineering.

In this paper, both Hooke's joint and the modern ball type constant velocity unit are considered. Methods by which the specific capacity and life endurance of a joint may be estimated are described. A method of calculation to select a suitable joint for automotive purposes is discussed.

Many lubricated machine contacts can be represented by cylindrical surfaces, and some of these contacts such as gears and rolling contact bearings have been studied extensively in recent years within the developing subject of elastohydrodynamic lubrication. When high loads are carried by the contact, side leakage is of little importance, but there is little evidence to indicate the quantitative importance of side leakage when the loads are light. This question has assumed greater importance with the recent recognition that many nominally parallel surface thrust bearings rely upon thermal distortion for their effective load-carrying capacity, since the deformed shape can be approximated by a pad of cylindrical form. In this paper solutions of the Reynolds equation are presented for the geometry of a cylinder near a plane lubricated by an isoviscous fluid. The resulting pressure distribution has been integrated to yield normal and tangential surface force components and the results are presented in the form of side-leakage factors to be used with the Martin solution for an infinite cylinder. The approximate ‘short-cylinder’ theory is presented and compared with the full computer solutions. The results are applied to two lubricated contact situations.

The analysis of the strength and stiffness of a diesel engine frame can be carried out in two parts; an analysis of the whole frame loaded by the inertia of the moving parts and an analysis of one line loaded by the explosion forces. This paper describes computer programs which have been used to make the above analyses in more detail and in less time than is possible by hand calculation. Results are given for an engine frame in which one program was used to determine the shear and bending forces on the frame and the loads on the ligaments between the various holes in the sides of the frame. Another program was used for the analysis under explosion forces; the walls were divided into a large number of elemental panels (each of side about 2 in) and the stress distribution (both direct and bending) in each of these elements was determined.

This paper describes investigations that were undertaken to identify the reasons for fatigue failures in cast iron and aluminium pistons of large diesel engines. The work involved experimental and theoretical stress analyses, and fatigue tests were made on specimens cut from actual pistons. It was found that four pistons which suffered fatigue failures had safety factors of 0·60, 0·68, 0·94, and 0·96, while three modified designs which performed satisfactorily were assessed to have safety factors of 1·1, 1·24, and 1·57. Special attention is drawn to the method of fatigue testing in which features typical of the piston are retained. While it is considered prudent to design for safety factors substantially in excess of unity, the results presented should promote confidence that this approach to the design of components is valid not only in this case but more generally.

Calculation procedures for the design of nitrided gear teeth commonly use a basic allowable stress which is constant for a particular core tonnage and case depth. The purpose of this paper is to determine the distribution of stresses in nitrided rolling elements, including gear teeth, and to illustrate the effects on the theoretical basic allowable stress of the size and geometry of the gears. Two criteria of fatigue failure are compared and the predicted values are compared with test results.

It is generally accepted that the performance of compensated externally pressurized journal bearings is superior to ordinary journal bearings. This paper presents the experimental results of performance studies of compensated and uncompensated journal bearings under static and alternating loading with an operating frequency of 327 c/min. The length/diameter ratio of the test bearings was 1/3 and both bearings had 10 closed-end grooves machined in the bearing surfaces. Film pressure surveys reveal considerable differences in the film pressure distribution in the two bearings, and explain the difference in the performance characteristics of these bearings. Comparisons are made in terms of measured load capacity, stiffness, and lubricant flow rates. The relative costs of manufacture and the operating costs of these bearings are also discussed.

This paper presents a theoretical and experimental assessment of the eight coefficients which describe the stiffness and damping properties of journal bearing oil films, together with their subsequent use in predicting the performance of a real rotor. The performance is checked experimentally.

The paper deals with large slow-speed marine engines with a section on crankshaft stresses, including dynamic torque calculation; static bending moments owing to firing pressure, inertia, and misalignment; combined torque and bending stress; and experimental determination of stress concentration factors in the fillet radii. Other sections deal with top end bearing loading including cyclic variation, piston head stresses, deflection of bearing housings, prestressing of main bolts, frame stiffness, and stress in combustion chamber studs as a result of pressure and thermal loading.

Dovetail pole root fixings, consolidated by tapered keys, are a feature of many large-diameter salient pole alternator rotors. This paper describes an investigation into the failure in service of such a fixing; the failure being ascribed to fretting fatigue and resulting from a combination of tensile centrifugal loadings and superimposed oscillatory loading. The failures were reproduced on a series of full-scale joint sections and the design of a special rig, capable of applying typical oscillatory forcing superimposed upon a steady tensile load of 100 tons, is described. The rig was used to assess quantitatively the effect of various measures taken to eliminate the cause of failure. The test specimens were hand fitted to normal manufacturing methods and, despite the inevitable variations in fit, failed in a consistent manner. The results of the tests provide a practical example of fretting fatigue, and are generally in line with the findings of a fundamental investigation using small specimens and published in 1967 by Field and Waters. The dramatic improvement resulting from relatively minor geometrical changes is clearly demonstrated.

To prevent or minimize the erosion of the last row rotor blades in large high-speed steam turbines it has been proposed to remove the film of water on the surfaces of the last row stator blades before it reaches the trailing edges and is swept off by the steam drag into the path of the rotor blades. The paper describes experiments with a cascade of hollow stator blades with various dispositions of slots providing communication from the blade surfaces to the hollow cavity. The same cascade was used in turn with three different wet air tunnels, the experiments thus covering a wide range of Mach number and Reynolds number. With suitably disposed slots, and bleeding a very small proportion of the working fluid, about 90 per cent of the deleterious water can be removed, thus preventing it from striking the moving blades.

The paper describes some apparatus which has been developed to measure those characteristics which are of interest in vehicle ride and handling. These are: Vertical, lateral and fore-and-aft tyre stiffnesses and damper characteristics. Suspension spring rates and friction, suspension geometry and suspension roll stiffnesses. Steering gear geometry and dynamic characteristics, moments and products of inertia of the vehicle and vibration characteristics of the seats. Some relevant theory and sample results are included in the Appendices.

Surface effects can influence the friction and wear characteristics and the fatigue behaviour of materials, so that any investigation of these phenomena necessitates careful study of the surfaces involved. Changes in topography, chemical nature, structure, and physical properties of surfaces subjected to sliding, rolling, abrasion, corrosion, deformation and repeated stressing have been studied using optical and electron microscopy and micro-probe analysis to obtain information of use in the elucidation of failure mechanisms and their controlling factors. Specific information from each technique enables comprehensive data to be accumulated. The paper reviews work which has been carried out in recent years, mainly from the technological viewpoint.

Several procedures for designing externally pressurized journal bearings have been presented. These vary from simple approximate solutions based on the bearing stiffness in the concentric position to full relaxation programs. In this paper the authors compare the predicted performance of bearings with experimental results. In addition to the static results obtained by the authors on a typical four-pocket grinding spindle bearing, the dynamic results of Patrick and Chen are also treated.

This paper develops a simple relationship between distance upstream and downstream of a row of blades and the velocity disturbances created by the passage of those blades. The relationship is compared with values obtained by numerical computation and measurement in an experimental axial flow compressor rig. The magnitude of vibration exciting forces and/or noise radiation sources due to potential flow interaction on a stationary row of blades is directly related to the magnitude of these velocity fluctuations. The analysis is therefore a basis ( a) for estimating the inter-row spacings for a turbomachine such that potential flow interactions are reduced below the levels of other forms of interaction, and ( b) of translating practical experience of noise and/or vibration excitation in existing machines to give reliable predictions for future designs. It is found that while potential flow interaction effects drop rapidly with increasing blade row spacing for low-speed machines, this is not the case for high speeds and very large spacings may, in fact, be required where the rotor blade speed is equivalent to a high value of Mach number. It is also found that the rate of decrease in effect is related to the circumferential wavelength of the disturbance, and comparisons between machines based on blade chords (as is the present normal practice) are meaningless unless the ratio pitch/chord is constant.

This paper describes an experimental investigation into the static performance of a floating-pad journal bearing with four pads. Up to moderate eccentricities the test bearing was found to have a somewhat greater load capacity than an equivalent plain journal bearing, and there was no evidence of instability. The geometry of the bearing also allows high lubricant flows to be used for cooling purposes without requiring excessive supply pressures. Some difficulties were experienced with starting under load and with a self-excited vibration which occurred at moderate to high eccentricities. The results of the investigation suggest that this form of bearing is most suitable for low-load, high-speed applications.

The dynamic bearing characteristics of plain seals having appreciable length-to-diameter ratio are analysed by a perturbation method giving linear results. Non-linear effects have been treated in a more precise analysis, and a summary of the method and comparison of results with the simpler linear results is given. Some experimental results are given and compared with theoretical predictions. It is shown that length-to-diameter ratio has a considerable effect on seals of appreciable length, e.g. balance pistons, and that dynamic components of bearing action due to shaft rotation are comparable with the purely hydrostatic centring forces. Experimental results are shown to be reasonably in agreement with theory. It is concluded that the linearized bearing coefficients of seals given in this paper form a sufficiently sound basis for use in the vibrations analysis of pump rotors.

This paper describes details of a solution to the finite journal bearing problem, including the effects of elastic distortions in the bearing liner. Solutions for the particular case of a bearing of length to diameter ratio of 1, outside to inside diameter ratio of 1·3, and Poisson's ratio of 0·4 have been obtained and the results are compared with experimental results obtained with a Delrin plastic bearing. Good agreement between theory and experiment has been obtained. The theory, which is applicable to soft shell bearings supported in a rigid housing, predicts that: (1) the ratio of the peak pressure to the average pressure is reduced; (2) the bearing stiffness is reduced; (3) the minimum film thickness for a given load is reduced by elastic distortions at eccentricity ratios less than unity; and (4) the friction is reduced.

Geometrical imperfections in gears can alter significantly the ways in which load and motion are transmitted. This paper describes these effects in a manner which is applicable to all types of gears, leading to a greater understanding of the problems of accurate indexing, vibration and noise, and composite inspection techniques. Theoretical and experimental comparisons are made between single and dual flank composite testing methods, the results of which suggest that these methods need to be reviewed and extended.

In this paper the choice of design and materials for artificial human joints is considered as an engineering design problem, and some possible solutions which result from this approach are discussed. Various types of plain bearing design are examined and the use of flexible connecting members is suggested as an alternative possibility. It is considered that the plain bearing designs must be based on the requirement that they should operate with a full fluid film of synovial fluid, and possible designs and materials are discussed from this point of view. The effect of fixation problems on the total design is also examined, and designs which appear to offer a rational solution are proposed. It is concluded that for bearing engineers to make any major contribution to the design problem requires more information on the variation of joint load in both magnitude and direction in relation to the timing of joint movement. Research into the performance of plastic on plastic bearings is also desirable with more information on which plastic materials are biologically acceptable.

Existing methods of stress analysis for pistons assume axial symmetry of form and loading; they cannot predict stresses in the region of the gudgeon pin bosses. In the method of analysis proposed in this paper a piston is idealized as a portal, including the gudgeon pin bosses; allowance is made for the restraining effect of the rest of the skirt on this portal; the deformations of piston and gudgeon pin are made to conform. The idealized problem may now be formulated for solution by a standard structural analysis program; however this involves a considerable amount of hand calculation, especially for the evaulation of thermal stresses. In order to reduce this, a special purpose program was written. This uses the force method of structural analysis and can evaluate both mechanical and thermal stresses. Stresses measured during rig tests are compared with those calculated using the new method. Good agreement is obtained and, in the authors' view, the use of the ‘restrained portal’ idealization is confirmed.

In order to achieve the highest reliability in operation of large hydraulic turbo-machines, it is essential, amongst other things, to determine the magnitude not only of the static but also of the dynamic stresses under all operating conditions. The adjustable guide vanes of reversible turbine-pumps, for example, form a particularly critical component from the point of view of dynamic stresses, especially during pumping operations. This paper deals with the possibilities of predicting resonance conditions and of obtaining the magnitude of dynamic stresses in full-size machines on the basis of tests on model machines. The laws of similarity for forced and self-excited vibrations as well as vibrations in resonance are derived. It is shown that all kinds of hydraulically excited vibrations can be examined by model tests. It is permissible to carry out these measurements under reduced test head, provided that for the tested components different materials are used for the model and full-scale machine, and that the head ratio model-prototype is chosen accordingly. Suitable materials for model tests are recommended. The evaluation of the dynamic stresses for the full-size machine, based on the measured data of model tests, is shown for resonant and non-resonant conditions. Further, as an example, vibration investigations are described on the movable guide vanes of a pump-turbine, that is, using the derived laws of similarity.

Information available for the calculation of stresses in and flexibility of mitred pipe bends (including lobster- back bends) under elastic conditions, for internal pressure loading, and in-plane and out-of-plane bending loading, is assessed. The results of Owen and Emmerson (4) for single mitre bends under internal pressure loading and their application to lobster-back bends are discussed. The stiffening effect of a ring of varying degrees of stiffness at the joint is considered in the light of some experimental results. The results of in-plane bending tests on two separate single mitre bends of different mitre angles are compared and a further comparison is made with published results on lobster-back bends. It is concluded that the results for single mitre bends under in-plane bending cannot be applied to lobster-back bends, which are more closely approximated by smooth-bend theory. Out-of-plane bending loading is briefly discussed.

Currently, systems which include diesel engines are designed and modified to avoid torsional vibration trouble on the basis of experience, and usually in the case of simpler systems a feasible one is found after few attempts. However, even these few attempts can be time consuming if the analysis is performed on a digital computer on a job shop basis. The purpose of the work described here, using a computer program to start from basic design data for the engines, alternators, and the like, is to achieve automatically a practically feasible and safe solution for the system, allowing for production restraints such as standard ranges of size of the various components. Economic considerations are also included. The systems examined are unbranched; simple examples being given.

A range of tests on simple radially grooved thrust washers is reported in which the number of grooves is varied from two to six. The mode of action by free thermal distortion is confirmed, the minimum film thickness occurring at 0·6–0·8 of the pad length. A degree of random waviness is shown to be beneficial, although it is likely that only half of the bearing circumferential length contributes to load carrying. The optimum number of pads appears to be four. The authors explore the use of multiple regression analysis in the design of such bearings. If several effects interact in the usual design process, it is shown that this technique can often give the direct influence of each design parameter separately without the need for iteration.

The design theory for transmission shafting, both in the elastic and plastic states, is reviewed, with consideration of the effects of splines, fillets, and similar stress-raisers. Data are presented on the actual strengths of such shafts for various steels under static and fatigue loading, and the nature of the fatigue phenomenon briefly discussed. Some typical shafting failures are described, and the appropriate lessons drawn.

Work so far carried out confirms that the dynamic variations of axle load are similar in magnitude to the changes of axle load due to quasi-static weight transfer. The dynamic responses of the suspension, particularly the primary stage, are therefore considered to be of appreciable significance so far as adhesion is concerned. Road test results obtained with Bo-Bo electric locomotives having spring and kinematic drives as well as axle-hung motors are presented and these indicate the effect of specific suspension arrangements on variations of axle load.

The pure braking and cornering forces of a given tyre on a given wet surface at a given speed are functions of braking slip and slip angle respectively. These functional relationships are of great importance in connection with the handling and stability of vehicles and the behaviour of brakes and antilocking devices. When cornering and braking occur simultaneously the forces interact and the functional relationships are modified, with consequent effects on vehicle characteristics. The empirical data available at the Laboratory on these relationships, both for pure forces and for forces in combination, are presented. The known effects of such factors as speed, road surface texture, tread pattern, tread resilience, and tyre construction are reported, but since the work was carried out at various times and for various purposes the coverage is incomplete and uneven. The test vehicle, test procedure and the methods of measuring and evaluating the data are briefly described. Some data have been analysed statistically to discover the dependence of the various frictional coefficients on factors such as those mentioned above. It is believed that greater use should be made of these statistical techniques in future work. Raw data can also be converted into such practical quantities as braking distances, wheel locking times, etc., by computing theoretically the behaviour of model braking systems. An example of such a study is given. The principal features of the mutual effect of braking and cornering forces may be summarized as follows: (a) If curves of sideway force against slip angle are plotted for constant values of braking force, there is little change of initial slope, but the maximum sideway force and the slip angle at which it occurs are reduced. If, however, the curves are plotted for constant braking slip, then the initial slope is decreased and the slip angle for maximum sideway force is increased; at this same angle the sideway force at zero braking slip will only be slightly greater. (b) The presence of a fixed amount of angular slip reduces the initial slope of the curve of braking force against braking slip, reduces the maximum value of the force, and increases the slip at which it is attained. (c) In a few cases examined, if resultant force is plotted against resultant slip for a given speed, tyre, and surface, there is a tendency for the points to lie in a fairly definite band having a form similar to that of the braking force–braking slip curve.

An experimental rig has been constructed to establish the vibration characteristics of plain, dammed groove, and tilting pad bearings having a variety of clearances. The rig is perhaps unusual in not being a small piece of laboratory equipment, but consisting of a naval anti-submarine frigate turbo rotor (15 000 rev/min) running in two journal bearings each supported by an isotropic cantilever system of variable stiffness. Two pairs of induction probes mounted close to the test bearing monitored the shaft movement and provided an oscilloscope display of the movement of the bearing centre. Detailed information about the shaft oscillations was obtained from a wave analyser and vibration meter. The static rig natural frequencies with various bearing housing support stiffnesses were primarily determined. From preliminary running tests a fairly stiff bearing housing support, which gave a strong vibration of approxi-mately half rotational speed, was chosen for all subsequent work. With the rotor wheel balanced the tilting pad bearing tested was found to be free from serious vibration. The dammed groove type of bearing tested was also stable at low speeds but was not capable of stable operation at 15 000 rev/min and above. The plain bearing tested produced serious vibration above 9000 rev/min. The frequencies and amplitudes of vibration which occurred in different bearings with a deliberate rotor unbalance were also measured. Increase of the clearance lowered the onset speed of serious vibration. The locus of the mean position of the journal centre of the tilting pad bearing was almost a vertical straight line whereas all the other bearings showed the usual approximate semi-circle. Instantaneous oil film pressures up to 900 lb/in2 were measured using a pressure transducer in the bearing. As the design particulars of the bearings tested differed somewhat the results were not strictly comparable but nevertheless an approximate design rule has been suggested. Practical data provided about the vibration behaviour of the different journal bearing designs can be used by the many theoretical investigators in this field. This is important in view of the lack of practical attested results to substantiate or otherwise the many theories advanced.

Work on subsonic compressors is reviewed and traditional British and American methods of design are compared. Transonic compressors are now well established and recent progress, demonstrating the necessity for properly shaped blades in machines having high tip speeds, is examined. Advances in the computer-aided design of turbomachinery are described: these include the ‘Matrix through-flow’ and ‘Streamline curvature’ methods of calculating the meridional flow. Experimental results give preliminary confirmation of a method of annulus wall boundary layer prediction. Finally, current and future developments are discussed and an attempt is made to describe how an axial flow compressor may be designed a decade hence.

The results are given from a comprehensive investigation of the exhaust pipe configurations for five-, six-, seven-and eight-cylinder turbocharged two-stroke engines. The tests were carried out under cold conditions on a multi-cylinder pulse generator. For each engine two firing orders and the maximum number of alternative pipe arrangements compatible with minimum interference between cylinders were examined. Three nozzle sizes were tested for each basic configuration. The test results were analysed to determine the maximum efficiency of the exhaust pipe, expressed as the availability factor, the average scavenge ratio and the distribution of retained air between cylinders. Results of the investigation showed that in the multi-cylinder system the scavenge ratio increased with increase in nozzle area ratio, increase in the number of cylinders discharging into the pipe system, and with decrease in pipe length. The gas retained in the cylinders was found to increase with decrease in nozzle area to pipe area ratio and decrease in pipe volume. The availability factor was found to increase with decrease in nozzle area to pipe area ratio and decrease in pipe volume.