CONSIDERATIONS ON BOGIEDESIGN WITH PARTICULAR REFERENCETO ELECTRICRAILWAYS

Paper presented to the Institution of Mechanical Engineers by W. S. GRAFF-BAKER* and reprinted in the Journd by kind permission of their Council.

PAPER No. 513 SYNOPSIS An examination is made of the dynamic characteristics of wheel sets and bogies, and of the various forces which act upon a bogie under service conditions. The fundamentals of bogie design are considered, and particular mention is made of recent developments in Ifiethods of body suspension. Problems of frame construction, braking, and power transmission are also considered. The paper concludes with a survey of the development of bogie design on the railways of London Transport Executive and elsewhere, and a restatement of the basic problems in the relation of bogie to track. INTRODUCTION Railway passenger rolling stock was originally constructed as four-wheeled vehicles of limited length carried on two ayles. The requirement of a longer body in due course led to the introL :tion of a third axle fitted intermediately between the other two, and this construction, with detailed variations, was extended to the point where it could not be taken further. The demand for still longer bodies, however, persisted and this requirement could only be met by the provision of pivoted bogies under the ends of the car, each bogie in the first instance being

* Late Chief Mechanical Engineer (Railways) , London Transport Executiv?, Past President, Institution of 1,ocomotive Engineers. 306 BOGIE DESIGN WITH REFERENCE TO ELEC*RIC RAILWAYS 307 provided with two axles. By this means a long body could be provided, able to negotiate the curves normally found on a railway, with the wheels separated by a limited rigid wheelbase. The functions of a bogie can, therefore, be described as providing running gear for a long-bodied vehicle while' keeping the rigid wheelbase within relatively small limits and providing comfortable riding for the passengers.

CONDITIONS OF OPERATION Before considering the design of bogies calculated to provide comfortable riding, the factors governing their construction and operation must be considered in some detail. These can be summarized simply as :-

(a) Use of " coned " wheels. (b) Use of wheels mounted solid on the axles. (c) Irregularities of the track on which the vehicle runs, divided into vertical and lateral irregularities. (d) Provision of means of acceleration. (e) Provision of means of braking. (f) Minimum maintenance cost.

(a) Coned Wheels. The normal practice on railways is to use wheels coned on the tread.at 1 in 20 so that the tendency of the wheel set shall be to take a mid-position on a straight track-or such position as will equalise the tread diameters in contact-and so that the wheel set, if deflected owing to centrifugal force on a curve, will run on two different diameters of tread, thereby compensating to such extent as is practicable for the curvature (Fig. 1). It has been demonstrated that such a set of coned wheels will take a sinusoidal course when travelling along a straight track; that is to say, it will inherently travel from side to side when progressing along a track. All studies of this fact have necessarily been based upon wheels and track both in new condition. In practice, neither wheels nor track are more than instantaneously in new condition and, while the sinusoidal motion is fundamental, owing to variations in the practical condition the motion is far more irregular and extensive than is indicated by theory. It might be possible to predict with any given wheel set and rail condition the path which would be followed but, owing to the large number of probable variables, all that can be said is that in practice there is an effective tendency of the wheel set to move from side to side irregularly rather than regularly. So much so that experiments have been made with cylindrical-tread wheels with a view to checking the side move- ment. This is at the expense of the somewhat equivocal benefit of the compensation for CUN~S by coning. As, however, it is relatively difficult in the type of lathe used to turn accurately alike such large diameters, ccning of 1 in 100 had to be adopted in practice and has been used on a considerable scale. Whilst this results in improved running, the treads require more frequent turning to maintain the effect. 308 JOURNAL OF THE INST. OF LOCO. ENGINEERS

(b) Rigid Wheel Sets. What has been said about the sinusoidal travel of a wheel set is related to sets in which the wheels are rigidly attached to the axles, and it can be assumed that the path followed by a set having wheels free on the axles could well be substantially different from that of the rigid wheel set. The question is, however, only of academic interest since, for reasons of braking, it is essential that the wheels should be rigidly attached to the axle. This requirement arises from the consideration that maximum braking is required often in service and inevitably under emergency conditions. Such braking is commonly achieved by pressing brake blockson to the wheel tyres at the maximum pressure which will avoid the wheels skidding on the rails.

OUTER WHEEL INNER WHEEL xD. Circumfesplce of inner wheel at contact diameter. x(D+ 0.0688). Circumference of outer wheel at contact diameter 0.216 in. Circumferential difference at contact diameters.

FIG.1. CONED WHEELS: WHEEL DIAMETERDIFFERENTIAL ON CURVEDTRACK

To obtain the maximum braking, it is necessary, or at least convenient, that the pressures applied to each block on both wheels of the set shall be equal, as well as maximum. Under conditions of railway operation, it is improbable that the superimposed weight on the axle will be distributed equally on the two wheels of a set .at any one moment; for example, in going, round a curve, the centrifugal force will cause the reaction on the outer wheel to be greater than that on the inner. In this case, if the wheels were free on the axle and the maximum force (based on equal bearing) were applied to the brake shoes on the inner wheel, where the weight has been relieved by centrifugal action, this wheel would inevitably “pick up” and skid on BOGIE IIESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 309

the rail. Conversely the weight on the outer wheel would be sufficient to permit even higher brake-block pressure on that wheel. The attachment of the two wheels rigidly to the axle ensures that the total friction force of the brake shoes is applied to the total weight carried by the wheels on the rails, without reference to the distribution of the weight on the two wheels. If wheels free on the axle were employed it would be necessary to limit the braking on both wheels to a figure which would prevent either wheel from picking up under the worst conditions of lateral weight transfer likely to be met in service owing to centrifugal or other effects. The result would be inadequate braking.

(c) Truck Irregularities. Track irregularities may be divided into tno kinds, vertical and horizontal. Vertical irregularities are constituted by the conditions arising from wear of the rail head, unevenness at rail joints and the vertical variations in the travel of the wheel caused when traversing points and crossings. Lateral’ irregularities arise from side cutting of the track due to lateral oscillation of the wheel sets, whereby the flange of the wheel is caused to act as a cutter on the rail, since its reaction on the rail is of a sliding, rather than a rolling, character. In any case, play has to be allowed between the flange of the wheel and the rail, to provide working clearance and to permit coning to be effective in adjusting the contact diameters of the wheels to suit the curvature of the track or to equalise the diameters in contact on straight track, and this has been shown to cause side movement.

(d) Acceleration. On electric railways, with motors mounted in the bogies, it is required that a motored bogie should not only carry , the motors but also be able to transmit the forces developed to accelerate the train and maintain it in motion, this force being derived from the adhesion of the wheel on the rail when the axle is caused to revolve by means of the motor.

(e) Braking. Braking is an extension to all bogies of the requirement indicated for acceleration, in that the braking force for the whole weight of the train must be provided on the bogies up to the limit of the adhesion between the wheel and the rail. Railways are, generally speaking, electrified to provide a more intense service of trains. Obviously, the requirements of acceleration and braking become of paramount importance in the provision of services of high intensity, since not only do high acceleration and braking provide a greater capacity for the line, in trains per hour, but also they permit the operation of a given schedule most economically, in regard to energy consumption, by avoiding the need to run at high speed. In some cases, it has been the practice to motor every axle on a train, not only to provide the maximum acceleration but also to enable the use of electro-dynamic braking and so to avoid the use of brake shoes, which, after all, may seem to constitute a somewhat crude, if simple, method of dissipating the kinetic energy stored in a moving train. 310 JOURNAL OF THE INST. OF LOCO. ENGINEERS

Electric braking can be regenerative or rheostatic. The kinetic energy of the moving train is reconverted to electric energy, in the former for use elsewhere, and in the latter to be dissipated as heat in resistances. In each case the traction motors are used as generators. The economics of regeneration depend on the cost of energy, the frequency of stops, and the speed at which braking is applied. In either case, for maximum braking, every axle must be motored or brake blocks must be applied on the unmotored axles. If traction motors are to be used for braking as well as for propulsion they must be larger and heavier, or more numerous, than if needed only for the latter purpose. An economic comparison of electrical and brake shoe methods of braking, taking into account not only running and maintenance costs but also interest and depreciation charges, shows that to motor every axle on the train is seldom economic. It is probable that the economic limit for acceleration is reached by motoring, at most, half the axles. It is, however, essential to brake every wheel on a train and, despite its crudity, the provision of brake shoes to every axle can be shown to be the most economical way of dealing with the problem, except where regeneration can be effectively used as, for instance, where energy is expensive, stops are frequent, or gradients are severe or long. I Where winters are severe the availability of dynamic braking current for car heating is a valuable factor, particularly for rapid transit systems, and reduction of brake shoe dust is a feature of dynamic braking especially attractive in subway work when non- metallic blocks are not available. In emergencies in intensive services operation at the maximum possible braking is necessary, that is, at a level limited only by the adhesion between wheel and rail, which amounts to a deceleration of 0.2-0.25 gravity.

EFFECTS OF CONDITIONS OF OPERATION The effects of the requirements and conditions of operation on the design of a bogie will now be considered.

Coned Wheels The lateral movement of coned wheels requires that the bogie shall be so constructed as to permit that effect without transmitting the motion to the car, body in which the passengers are seafed. This requirement must be met not only in respect of the theoretical sinusoidal movement of the wheel sets, but also in respect of the irregular lateral movements resulting. from wear and tear of the wheels themselves and of the track.

Rigid Wheel Sets The effect of these is more to simplify the bogie design than otherwise, and to permit the provision of brake-gear bearing equally on the wheels on both sides of the bogie. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 311

Track Irregularities The vertical irregularities of the track must be absorbed by the bogie without being transmitted to the body and, in so far as vertical irregularities of the track may cause a tendency of the vehicle to roll, this rolling must be controlled by the mechanism of the bogies. The lateral track irregularities add considerably to the effect of the natural motion of the wheel sets. Acceleration The bogie must at all times ,be able successfully to transmit the tractive effort of the wheels through its frame and the body to the train as a whole. Braking The bogie must at all times be able to transmit the forces required to stop the train.

CENTRE PIVOT

a Spring-plank

b b Bar-suspended FIG.2. BOLSTERSUSPENSIONS GENERAL CONSTRUCTION The normal design of a bogie, as the result of considerable experience, consists of a frame mounted on two or three axles. These are provided with axle-boxes which can slide vertically within the horn-guides of the frame, the load of the vehicle being transmitted from the bogie frame to the axle-boxes by means of springs. The bogie frame is of rectangular open-frame construction, provided in the 312 JOURNAL OF 1HE INST. OF LOCO. ENGINEERS BOGIE DESIGN WITH REFERENCE TO ELECTRI(' RAILWAYS 313 314 JOURNAL OF THE INST. OF LOCO. ENGINEERS

middle with two transverse members or transoms, between which is mounted a bolster provided with means for lateral movement. The car body rests on the bolster, with a king-pin to relate the bogie bolster position to the body bolster. Sliding or equivalent surfaces are provided at the ends of the bolster so that the car body is related laterally to the bogie bolster whilst permitting the bolster and the bogie on which it is mounted to turn in a horizontal plane. The bolster is supported by springs which generally rest upon a spring plank or tray suspended from the transoms by links or hangers, so that the bolster can move vertically within the bogie frame. under spring deflection and travel laterally within the frame by virtue of the links (Fig. 2a). The bolster hangers are usually inclined to the vertical, so that the bolster, with its associated parts, tends to return to a central position after deflection. The restoring moment increases with the deflection, so that control on movement is afforded. In some cases double bolsters have been used-one suspended from the other. The - bolster being laterally rigid with the body, the flexible suspension of the bolster permits the wheel sets and bogie frame to move laterally without affecting the body. If three axles are used two bolsters must be provided, with their associated mechanism, one between each pair of axles, the bolsters being coupled, together for unified attachment to the car body.

Dean Suspension For a number of years, mainly on one British railway, bogies without bolsters were constructed, the car body being connected directly to the bogie frame by links connecting the top of the frame through springs to the lower end of a vertical post extending down- wards from the body (Fig. 3). Thus, the whole of the body became, in effect, a bolster on hangers. This form of contruction is extremely simple and has much to recommend it, but is found in practice to lead to less than optimum riding at speeds over about 50 to 55 miles per hour. This suspension has been adopted in Switzerland for some recent experimental pneumatic-tyred rail vehicles, the upper end of the hanger being connected to arms mounted on torsion-bar springs. It is also to be used on the Metropolitan-Vickers gas-turbine locomotive for British Railways.

BOGIE SPRING SYSTEMS It is found desirable in practice that the combined spring system of the bogie should be aperiodic to avoid setting up resonant movements. If all the springs are spiral this is somewhat difficult to achieve over the whole range of impulses, and generally results in the need for shock or recoil absorbers, a practice which is tending to grow in the United States. It has, therefore, been common practice to provide two sets of springs on a bogie, one consisting of spiral springs and the other of laminated springs, which have considerable damping effect. There are two main practices in regard to the arrangement of such ilifferentiated springs, which may be described as British and American BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 315

(Fig. 4). Normal British practice is to provide laminated half-elliptic springs over the axle-boxes, with spiral or rubber auxiliary springs at their ends, and spiral springs under the bolster. American practice is, normally, to provide laminated elliptic springs under the bolster. The bogie frame is supported by spiral springs on an equaliser bar resting on the two axle-boxes. Variants of these practices have occurred not only in Britain and the United States but also on the Continent, and the number of these variants may be regarded as some index of the feeling that the basic designs commonly used are less than perfect (Fig. 5).

a British typ; b American type FIG. 4. BRITISHAND AMERICANSTANDARD BOGIESPRING SYSTEMS

The bogies for the Stockholm underground railway are built with frames in which the axle-boxes are rubber-mounted and have little relative movement. The body load is taken on a central. table mounted on spiral springs, which are arranged to allow lateral move- ment as for a bolster. This bogie is related to modern American practice for underground lines and has rnotors mounted wholly on the bogie frame (Fig. 6). Some use has been made of torsion-bar springing by the Swiss (Fig. 7). In one application, a square torsion-bar runs along the side of the frame, being attached to the frame at the centre and connected to the axle-boxes through arms at the ends. Adjustment for bogie height is provided at the central fixing by means of an arm which can 316 JOURNAL OF THE INST. OF LOCO. ENGINEERS be adjusted to turn the torsion-bar by a screw abutting on the frame. The body is sometimes supported on side rollers coupled to the bogie frame through swing links and a second longitudinal torsion-bai- system (Fig. 7a). In current Swiss praktice, the axle-boxes are connected to the bogie frame by cylindrical guides surrounded by helical springs. The cylindrical guides are oil-filled and constitute a well-lubricated guiding surface as well as being constructed as dash-pots to control vertical oscillation of the springs. In this case the bolster may rest on the centres of two inverted half-elliptic springs arranged parallel to the side frame. The ends of these springs are suspended by links from the side frame, with freedom of lateral and longitudinal movement. In the latest construction the functions of this half-elliptic spring are taken over by two torsion-bars secured at the bolster centre and connected at their ends' to arms corresponding to the halves of the half-elliptic spring (Fig. 7b). In this case the bolster is much simplified since the whole weight of the body is taken at the ends through rubbing pads and the centre pivot is used solely for centring. Adjustment is provided at the fixing points. In either casr the vertical load distribu- tion on the frame is improved over normal practice by the bolster hangers being brought nearer to the axle-boxes instead of being concentrated at the centre of the frame. These bogie frames are, therefore, small in vertical dimension. They are, howewr, robustly designed to take lateral or racking loads. Comfortable vertical riding depends to no small extent on free- moving springs; that is to say, springs having the maximum permissible deflection per unit of loading. A set of bogies is running cxperifnentally on London Transport in which the bolster springs, spring plank, and hangers are replaced by a pair of rubber springs in shear. disposed at appropriate angles between an extended bolster and the outside of the bogie frame (Fig. 8). These springs are proportioned to give the same character- istics as'their more normal predecessor and the riding is the same. The springs deflect vertically to correspond with the bolster springs. and horizontally to correspond with the effects of the bolster hangers with their normal centring action. The advantages are the elimination of all wearing parts and a saving in weight at a capital and maintenance cost certainly not higher than normal. Further developments of the principle will be referred to later.

Control of Rdling A point arises as to the disposition of bolster springs. On thr, stiffness of these springs and the side bearing springs depends the riding of the car in the vertical plane. The bolster springs usuallj~ lie between the bogie side frames, and the side bearing springs lie outside the side frames. If the body tends to roll, the springs on one side must control the roll by their compression reaction. The nearer the springs are to the longitudinal axis of the bogie, the stiffer they must be to control rolling and the less contribution can they make towards good vertical riding. The side bearing springs are as far out as is practicable :I :I $1 PART SECTION ON C.L. SECTION THROUGH BOGIE il u/ U' LOOKING AT BOLSTER

SECTION AA

! (SIDE ROLLER OMITTED)

0 I 2 3 FEET 4 I I I Id -.I;- FIG.8. LONDONYRANSPORT EXPERIMENTAL RUBBER-MOUNTED BOLSTER BRAKE ‘OFF” BRAKE ‘ON”

TO EJECTOR TRAIN PIPE. - TRAIN PIPE. - II- ’ il1 1

GRESHAM’S DIRECT ADMISSION VALVE FOR VACUUM BRAKE. FIG.9 B:)GIE DESIGN WITH KEFEREXCE TO ELECTRIC KAIL\VA\’S 317

but the bolster springs as normally situated are inconveniently close to the centre-line, and have to be undesirably stiff if rolling is to be adequately checked. This has from time to time been recognised as, for instance, on certain dining-car bogies, where the bogie frames were set out beyond the axle-boxes to permit a longer bolster with springs disposed correspondingly farther outward. The axle-boxes were arranged in short, inne‘r side frames, each extending from the transom to the end bars. W. A. Agnew, chief mechanical engineer of London Transport, constructed t\vo bogies with the bolster and spring planks extended beyond the side frames; the spring hangers and bolster springs were mounted outside the side frames. The riding of the car with these bogies was improved ,in relation to similar cars bn the normal bogies, and this type of construction has been generally adoptcd in one form or another by London Transport since 1937. On surface line cars the bolster and spring plank were extended beyond the side frame (Fig. 9a); on tube cars the side frame was formed with a bulge at the centre to permit a wider than normal bolster spring-base (Fig. 96). Since each of these constructions has certain disadvantages a further development has been carried out for both surface and tube types of bogie. In this, the bolster spring is mounted on a bracket outside the side frame (Fig. 2b). Two bars extend through apertures in the bogie frame from spring to spring, and the bolster is suspended from these bars by hangers on knife-edges within the side frames. The riding of cars with such bogies of wide spring-base is steady as to rolling, while the easier springs thereby permitted give a soft ride in the class of service in which they run, both laden and unladen, in spite of thr high proportion of maximum load to tare weight of car.

WEARING PARTS It will have become obivious from the €unctions required of the various parts of the bogie that there will be a number of wearing parts. The wear and tear on bogies is accentuated by the heavy duty inherent in services calling for electric traction; the extreme case might be expected to be found on London Transport, where a train is required to stop and restart every five-eighths of a mile, on an average, and where it has been necessary to push the acceleration to the maximum economically, and the braking to the maximum physically practicable. The practice on this undertaking is to line all wearing parts with manganese steel, a material with considerable endurance which work-hardens and does not require an appreciable measure of lubrication for its successful operation. Such wearing plates are welded to the axle-boxes and horn guides, and to the bolster rubbing surfaces, which are provided to restrain the bolster from moving longitudinally in the bogie whilst permitting free vertical and lateral movement. Non-metallic liners are now bging developed and have been standardised in Italy. These promise well, but lack the possibility of attachment by welding. The design of bolster swing-link hangers offers some difficulty in regard to wear, since space considerations generally preclude the 318 JOURNAL OF THE INST. OF LOCO. ENGINEERS provision of pins with generous wearing surfaces. Furthermore, the lubrication of such pins is difficult as these bearings not only operate through small arcs under heavy loads, but also are subject to the deposition of grit and dirt from the track: Use of rounded knife- edges for the hanger system working in grooved mating parts avoids the need for lubricant and reduces wear and 'tear, and this practice has been adopted on British Railways and on the latest London Transport stock. Some developments have been made in the United State? towards eliminating all rubbing, and therefore all iearing, surfaces. Use is made of radius bars connecting the moving part to the fixed part, with space between these parts, and the movement is sufficiently rectilinear through the range required. The radius bars are of round section and coupled through heavy rubber washers to the parts to be related. Similar tendencies are beginning to show themselves in continental practice and in Britain by the use of links with " Silentbloc " type of bushes.

WHEELS AND AXLES The design of wheels and axles has been the subject of many papers including one by Spencer (1945)* and will not be dealt with extensively here. It may, however, be recofded that in Switzerland there are in operation, under heavy braking conditions, wheels having forged.. aluminium-alloy wheel centres with a view to reducing unsprung weight. The Reuleaux formula normally used to calculate axle diameter may not, in its usual form, be correct. A large number of axles running under tube cars with a low centre of gravity have not broken although considerably overstressed according to the Reuleaux calculation. A corrected formula, taking into account the height of the centre of gravity and making less improbable assumptions than the original formula, produces results similar to the original when used to determine axle sizes for normal railway rolling stock, and also conforms with the above-mentioned observations on tube axles. There is some, but not conclusive, evidence to show that motored axles tend to be more subject to crack development than are trailing axles, even at lower calculated stress values. This is perhaps due to the increased impact of the wheel flange on the rail when the side movement of a heavy motor, as well as of the wheel set and bogie, are being arrested. There is also evidence, again inconclusive, that the tendency to crack is less for axles the wheels of which are braked with non-metallic blocks, owing perhaps to reduced flange wear a?d lower adhesion values.

Plneumatic Tyres Experiments have been proceeding for a long time on the Continent with pneumatic rubber-tyred wheels for railway bogies.

* An alphabetical list of rpferences is given in the Appendix. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 319

For operation on normal railway track the wheel load must be limited to 1.2 tons. Consequently, the lightest bodies must be used, and the number of axles be increased to four or five per bogie. No overload capacity can be provided, and this makes the proposition of doubtful value for suburban services. While the bogie springing can be more simple than usual, some vertical vibration is experienced at certain speeds. Precautions are necessary in relation to liability to punctures. The reduction in noise, however, is remarkable. The economics of large numbers of rubber tyres compared with a few steel tyres seem to be unfavourable unless a supplementary fare can be charged. However, a wider running rail now under considera- tion might change the economics considerably. Further experiments are being made on the Paris Metropolitan railway with four-wheeled pneumatic-tyred bogies gunning on a wide, flat track- (Fig. 10). BOGIE FRAME DEFLECTIONS Deflections in bogie frames arise from the forces engendered in the bogie when it is passing round curves. Curvatures will occur which cannot be compensated by the coning of the wheels. In such cases three wheels of each four-wheel bogie must slip either longitudinally or laterally in going round the curve. Slipping is resisted by the adhesion between the wheel and the rail, so thht the tendency is for the axle to take an attitude other than radial to the curve. It can take this attitude only by deflecting or winding-up the bogie frame until the frame’s reaction is sufficient to overcome the adhesion between wheel and rail and permit the former to slip. The deflection which can occur must clearly depend on the built-in stiffness of the bogie frame. which, however, is usually an open rectangular frame with no inherent geometrical stability of shape. The process of winding-up the bogie frame until the wheel slips will be repeated as long as the bogie is traversing the curve. The effect of this repeated straining and de-straining may be to cause fatigue failures in the bogie frame. It can be shown experimentally and by calculation that the forces applied to a bogie from this cause are three times that of the whole load carried on the wheels multiplied by the coefficient of adhesion of wheel on rail, and may be regarded as applied longitudinally in opposition at two diagonally opposite corners of the bogie frame. With a fully loaded fifty-foot car of,the London Transport type, with non-metallic brake blocks, this may amount to four tons. With cast-iron brake blocks, coefficient of adhesion is increased by 25 per cent, which results in a distorting force of five, instead of four, tons. It is therefore necessary that the bogie frame, more particularly where curvatures are frequent or severe, should be made as stiff as reasonably possible whilst retaining a rectangular open structure, or be made inherently stiff by diagonal bracing in one or other (or both) of the bays of the bogie. BOGIE FRAMES Bogie frames are usually made of mild steel or steel castings. Higher tensile steels have also been used. 320 JOIIKNAL OF THE INST OF LOCO. ENGINEERS

Cast-steel bogie frames can be cast in one piece and can be provided integrally with lugs or projections required for any purpose, such as brake rigging, spring seats, etc. Little machining is required and such frames tend to be stiffer and somewhat heavier than other constructions. Mild-steel bogies are constructed with sections and plates or from pressings or any combination of these forms joined by riveting or welding, either acetylene or electric. .4 riveted bogie usually has plate side frames reinforced with angles at the edges, and has rolled sections or pressings for the cross-members, gussets being used at joints whrre necessary. Welded bogies have generally been designed along the same lines, with a plate side frame stiffened with flanges joined to the plate by welding. Similarly built-up sections or rolled sections may he used for cross-members and welded to the side frames. In modern continental practice there is some tendency toward5 welded framts uith box sections. In either welded or riveted construction, all brackets or othcxr projections are added to the main structure either by riveting, bolting, or welding as may be found convenient. If a welded construction is considered only as a structure to carry the superimposed load, it will be found that it lacks stiffness in a horizontal plane,. compared with a riveted frame. This may lead to the distortion due to track curvature unless diagonal bracing can be employed. Some care requires to be taken in welding structuxcs of this sort, subject to alternating stress, since any stress locked up in the frame will cause a predisposition to fatigue failure. The safest. way would seem to be to stress-relieve the complete frame after welding. Frames not so relieved have been found to open nearly an inch when cut through the end bar, while stress-relieving eliminated this. Although the niceties of welding are not within the scope' of this paper, it is suggested that welded gussets at corners should be welded along the neutral axis of the flange to which they are joined, and thit it should !e a required practice that electric welds should never be carried to a stressed edge of any structure of this kind. It has been observed that nearly all bogie frame fractures, whether in welded or riveted frames, may be attributed to stresses set up by frame distortion and generally arise at highly stressed edges. The current standard for construction of London Transport bogici (Figs. Ila and llb) is:- (a) A welded plate and rolled-section side frame, using channel- section flanges symmetrically disposed on the plate edge so that the flange members provide good lateral stiffness with low-stressed edges. This frame is stress-relieved. (b) Cross members riveted to the side frames with angle cleats and gussets. (c) Diagonal bracing (where possible) in one bay of the frame. (d) No welds reaching edges stressed in any plane. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 321

FIG.5. 1942 MILWAUKEERAILROAD BOGIE

FIG.6. BOGIEFOR STOCKHOLMUNDERGROUND RAILWAY

a Bolster and side springing with longitudinal torsion bars

b Lateral bolster torsion bars with helical ‘axle-box springs FIG. 7. SWISSDEVELOPMENTS IN TORSION-BARSPRINGING 322 JOURNAL OF THE INST. OF LOCO. ENGINEERS

a Surface line

b Tube line FIG.9. WIDE SPRING-BASEBOGIES

FIG. 10. PNEUMATIC-TYREDBOGIE, PARIS METROPOLITAN

b Bogie for " R47 " stock FIG. 11. CURRENTLONDON TRANSPORT BOGIE FRAME CONSTRUCTION BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 323

FIG.11. CURRENTLONDON TRANSPORT BOGIE FRAME CONSTRUCTION

BRAKING Braking of passenger vehicles is normally carried out by develop- ment of a force at a suitable point, by means of compressed air or vacuum, and transmission of this force to brake blocks which bear against the treads of the wheels. A brake cylinder is normally situated on the car body and is assodiated with a lever system from which a pull-rod extends to each bogie, whereon is mounted a further lever s stem the ultimate reaction points of which are the brake blocks. $he whole of this mechanism tends to be of a primitive character. It is virtually impossible to provide for adequate lubrication, while !he mechanism is not only exposed to the deposition of grit from the track but also is subject to considerable vibration during running, more particularly on the bogie. The efficiency of such a system is relatively low, while the forces involved are comparatively large. It will be appreciated that these forces in the bogie rigging cause reactions on the bogie structure. Wear on bogie rigging is related to the vibration when running idle rather than to the effects of use. It is common practice in the United States to fit the brake cylinders on the bogie, which to some extent reduces the amount of noisy and exposed mechanism by eliminating the body brake- rigging. A further step was taken in 1937 on the rolling stock of London Transport by the development of an individual brake unit (Fig. 12). This consists essentially of a brake cylinder with a single lever operating through a slack-adjuster directly on to one brake block. Eight brake blocks on a bogie, with the consequent eight brake cylinders, are not found in practice to cause difficulty. The reaction forces on the bogie are conveniently disposed, while the lever system is mounted completely in the unit case, and can be kept lubricated at all times. The mechanical efficiency of such a unit is high. While the first cost of such-an equipment is somewhat more than that of 324 JOURNAL O* THE INST. OF LOCO. ENGINEERS

a Diagram of unit

b Assembly perspective FIG.12. BRAKECYLINDER UNITS conventional brake-rigging , the maintenance costs are lo^ and the rattling associated with conventional brake-rigging is eliminated. The forces required to be developed depend on the material of the brake block. The normal brake block is of cast-iron but an altern a t'IVC exists in non-metallic form which can, if desired, have a higher coefficient of friction than cast-iron and, therefore, require less pressure to provide the same braking. An axle travelling at 50 m.p:h., with a 15-ton axle loading, may br associated with almost 3 million foot- pounds- of kinetic energy. To brake this axle at a retardation of 24 m.p.h. per second will involve an initial kinetic energy dissipation of about 500 h.p., or 125 h.p. per block if two brake blocks arc fitted BOGIE DESIGN WITH KEFERENCE TO ELECTRIC RAILWAYS 325 per wheel. The use of non-mefallic blocks involves a careful selection of tyre steel to avoid thermal cracking of the surface of the tyre, which is liable to occur with non-metallic brake blocks owing to the temperature of, say, 900 deg. C. (1,650 deg. F.) generated in the tyre surface under the brake blocks. This cracking also tends to occur with .cast-iron brake blocks, but generally the abrasive action of the blocks removes incipient cracks as soon as they are formed. With cast-iron blocks, not only is tread wear experienced but also flange wear. The non-metallic brake block causes less wheel wear since it is less abrasive; such wear as tak,es place is on the tread of the wheel and not on the flange, so that the tyre profile can he restored to standard with the minimum removal of metal. It is to be noted that the effective coefficient of friction, or adhesion, between wheel and rail is lower by some 20 per cent for a non-metallic blocked wheel than for a cast-iron blocked wheel. While the tyre surface is less smooth with cast-iron than the polished surface produced with non-metallic blocks, this difference cannot account entirely for the variation in coefficient, and it seems probable that the detritus from the non-metallic block may act as a lubricant. Where an axle is motored, it is possible to use the motors to provide dynamic or regenerative braking and avoid the use of brake blocks except for emergencies, but this is not necessarily economic.

AXLE-BOXES Axle-boxes house the bearings carrying the weight of the car and bogie on the axle. These bearings may be of the white metal type or of the roller-bearing type. The latter now tends to be more freely used. For services with very heavy rates of braking and acceleration there is no doubt that the roller-bearing axle-box is more reliable and, in the long run, less costly than the white-metal-bearing axle-box. The roller-bearing axle-box can take end thrust on the axle either through the same parts which carry the load, or it can be provided with a separate thrust bearing, the main bearing carrying only the vertical load. Spherical and tapered bearings fall into the former class, and plain cylindrical bearings into the latter. A simple form of roller-bearing box of the cylindrical type is used on London Transport, with a non-metallic end-thrust pad supported on the lid of the box and provide$ with shim adjustment to maintain correct running clearances (Fig. 13). Since wheel turning is frequent in such service, this form of box offers considerable advantages, as it can be- removed directly from the end of the axle without dismantling any part of the box or bearing, and the axles can then be put straight into the wheel lathe, centred in collets. Clearances between the axle-boxes and the bogie frame must be provided and, in the interests of good riding, these clearances must be maintained either by periodic repairs or by provision of non-wearing rubbing surfaces, or both. That a roller-bearing axle-box (except for the spherical type) is rigid with the axle, except rotationally, is important. If deflection of the bogie frame takes place by the displacement of one side frame in relation to the other, the axle (with its assembled axle-boxes) must 326 JOURNAL OF THE INST. OF LOCO. ENGINEERS

/.'I-- /.'I-- I I I I I --I-. I I I I I .-.d..

FIG. 13. LONDONTR~SPORT KOLLER-BEARING AXLE-BOX take an angle other than a right angle to the side-frame. If the deflection of the bogie frame is sufficient and the clearance between ihe axle-box and the horn-guides is that normally provided with plain bearings, the clearances between the axle-box and horn-guide may well be completely taken up across comers by the angular movement of the axle relative to the side frame. With further frame deflection, the axle-box and axle constitute a prise tending to force apart the horn-guides. This may give rise to bogie frame fractures or excessive wear. With plain bearings a certain amount of play exists between the bearing and the inside of the axle-box, so that the box can to some slight extent take up an angular position in relation to the axle. The same deflection of bogie frame will not cause this prising action on plain-bearing bogies as soon as it does on roller- bearing bogies with the same horn-guide clearances. Therefore, with roller-bearing axle-boxes more clearance must be provided between the horn-guides and the axle-box than with plain bearings, or measures must be taken to prevent deflection of the bogie frame. With spherical roller-bearings an angular movement of the axle-box in relation to the axle can occur.

MOTORS AND SUSPENSION The normal method of carrying electric motors on bogies is by nose-suspension (Fig. 14); that is to say, the motor is supported at one side by the axle and at the other by the bogie transom. At the axle side, the motor has two " suspension bearings," usually with split bushes, which embrace the axle, which is of a diameter suitable BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 327 A

f?I:, 328 JOIJRNAL OF THE INST. OF LOCO. ENGINEERS to provide a bearing area. At the transom side, the, motor has ;L projecting " nose " which rests in a pocket provided on the transom. Little change on any scale has occurred in this method for many years. Troubles with plain suspension bearings led London Transport to substitute roller-bearings, which lie at the ends of a tube surrounding the axle (Fig. 15). The motor suspension-bearing housings clamp on to the tube. With plain suspension bearings, the wear which arises allows the gears to move outyards from the cofrect centre-distancc-- a movement which is increased with plain motor bearings. As ;i result, it has been necessary to fill gear-cases with a bituminous typc of compound to prevent noise and to cushion the shock of pinion and gear teeth mating outside their pitch circles. Even so, wear of gears is considerable. With roller-bearing suspension and roller-bearing motors the centre distance is maintained, so a true lubricating greasc may be used, and wear on the gear teeth is negligible. In some cases the motor nose is supported from thr transom by springs or rubber, but experience would seem to indicate little, if any, advantage from the additional complication. A nose-suspended motor obviously has half its weight carried directly on the axle and, therefore, unsprung. The other half is sprung through the bogie side-bearing spring system. The unsprung part of the weight of a traction motor (weighing up to three tons) is obviously detrimental to the track. For this reason, over a considerable period, efforts have been made to carry the whole of the motor weight on springs. Where the wheel diameter is considerable-as on locomotives -it is posgible to surround the axle with a tube or " quill " connected Bt the ends through springs to the wheels: On this quill run the suspension bearings and mounted thereto is the gear wheel. On bogies and motors for passenger stock, room is .seldom, if ever, available for a quill. . Latterly there has been a tendency to use more and smaller motors. It has been found possible, for instance, to provide a flexible coupling between the motor pinion and the motor, and to support the motor wholly from the sprung part of the bogie, the gear case and gearing being supported only from the axle. The adoption on any scale of sugh systems on, for instance, an electric railway with a considerable fleet of cars may introduce complication and maintenancr out of proportion to the benefits to be gained. Unless a very simple form of construction with a minimum of non-wearing parts can be devised, the benefit of multiple small axle- hung motors in reducing unsprung weight per axle may be as much as can be expected in this direction.+

WEIGHT TRANSFER A bogie has not only to carry the body but also to stop it aritl-- for electric trains-start it. Reference has been made to the need for maximum acceleration and deceleration of electric trains. These depend on the adhesion of wheel to rail and this, in turn, on the weight carried at the relevant time by the wheel on the rail. It is desirable to motor an axle so that the driving torque developed approximates to the value of the BOGIE DESIGS WITH REFERENCE 1’0 ELECTRIC RAILWAYS 929 adhesion multiplied 6y the wheel radius. It is essential for maximum braking to brake a wheel so that the braking torque approximates to the adhesion multiplied by the wheel radius. Both maximum motor torque and maximum braking torque are, therefore, fixed quantities. Statically, the adhesion multiplied by the wheel radius is also fixed for any distribution of axle loading. Dynamically, however, conditions change. For instance, bodies are connected to ihe top of the bogies so that, when the mass of the body is decelerated by tl retardation of the bogie, its momentum causes a turning moment a1 ut the bogie-supporting bearings, throwing more weight on the le; ling axle and relieving the trailing axle of weight. Again, the braking and starting torques generated in the bogie must result in moments which alter the weight distribution on the wheels. These are only two Qf the many possible causes of changes of weight distribution. The changes in weight distribution alter or reverse with the direction of motion. Maximum braking or acceleration values can, therefore, be adjusted only to the minimum adhesion values after allowance has been made for weight distribution changes. These values must always be less than the static values, allowing for both directions of motion. A full treatment of this rnattri has already been published (Broadbent, 1950), and it will be seen from the considerations discussed therein that in some circumstances advantage is to be gained by building a bogie with the axles disposed at different distances from the bogie centre, and with onl) one traction motor. It is also clear that connecting the body to the bogie so that longitudinal forces are transmitted between them at the level-*ofthe axle centre-line produces substantially less weight transfer from axle to axle than does the normal construction. This feature has been adopted in the design of the P.C.C. tramcar in the United States (Fig. 16), which has a rigid downwardly projecting pillar attached to the body, reaching to axle level in the bogie. The drive for braking or acceleration is taken at the lower end of this pillar, and no moment is set up in the bogie itself to disturb maximum braking or acceleration torques.

BOGIE DESIGN ON LONDON TRANSPORT RAILWAYS The foregoing has constituted a general survey of bogie practice for electric railways as it stands today? and has endeavoured to indicate some of the factors determining bogie design. The whole question is in a state of flux. This applies even to electric locomotives. Figs. 17, 18, 19 and 20 show the trend of developments which have taken place in London Transport over the years in electric rolling stock for intense services.

Surface Line Rodling Stock In the first place, the bogies for the District Line were of American design, using cast-steel frames for motored bogies (Fig. 18a) and built-up structures for trailer bogies (Fig. 17a), following the designs for the Chicago Elevated, with American type of springing and outside-hung brake blocks, one per wheel. When further bogies were required, a form was used in line with British main line practice, 330 JOURNAL OF THE INST. OF LOCO. ENGINEERS with a pressed-steel motor bogie (Fig. 18b) and a structural-steel trailer bogie, with British methods of springing and inside-hung brake blocks, again, however, one per wheel. A later trailer bogie had a pressed-steel frame, but was otherwise as before (Fig. 17b), and, later still, bogies were used having built-up plate and angle side-frames and double,block brakes, but retaining conventional forms (gigs. 17c and 18c and d). At the next stage, an all-welded bogie was put into service with double-block rigging and bogie-mounted brake cylinders (Fig. 94. Motor and trailer bogies were the same, apart from springing, and only one motor was mounted in each bogie on motored cars. The brake-rigging was separate for each axle, since in some applications the motor was used for regenerative electric braking, and the brake blocks were only applied to the motored wheels at low speeds or in the event of failure to regenerate. On these bogies the bolsters were extended beyond the frames, with outside hangers to increase the stability of the body against rolling. The latest form of bogie has a welded side frame, with channel flanges and riveted cross-members. In these bogies brake rigging is dispensed with and individual brake cylinders are used.. The bolster springs are outside the frames, for stability, and they support crms- bars from which the bolsters are hung by inside hangers (Fig. llb). In all these designs, normal rubbing surfaces are provided to maintain correct relation between moving parts. Futpre developments are likely to eliminate rubbing surfaces and metal springs by the use of (rubber in shear. Tube LhRdling Stock Up to 1937, the development of motor bogies for tube cars followed fairly closely some stages of the process given above (Fig. 20). The trailer bogie on tube cars has always been a problem in design on account of the low car-floor height. The earliest designs had a frame suspended from springs over the axle-boxes, with swing bolsters (Fig. 19a), except on the early Central Line cars which had pressed- steel frames and sprung, but not swung, bolsters (Fig. 19b). Later designs followed the Central Line bogie, with the addition of a swing bolster (Fig. 19c), and the Central.Line bogies were modified to this pattern. All these had single brake-blocks. In 1937 it was decided to abandon the existing practice of having large motors on a few cars in a train, and to drive a larger proportion of the axles, using smaller motors, and a bogie was produced, suitable for both motored and trailing positions, of low height and carrying only one motor (Fig. 9b). These bogies are all-welded and have side frames opened out at the centre to admit a long bolster with widely-spaced springs. Braking is by two blocks per wheel, with individual brake cylinder units. Further developments for tube bogies aim at the reduction of wearing surfaces, as in the case of the surface bogies (Fig. 2Oc).

GENERAL It will be seen that the course followed in design has been in the direction of better riding, better braking, and lower maintenance costs. Riding is governed by springing and its disposition. The wide spring- FIG. 16. Ixnv THRUST-LINEBOGIE : P.C.C. TYPE(UNITED STATES)

LZ ISqualisrr bar, type K ' b I'rcssed steel, type c c Wclded steel, Mark 11, 1949 type FIG !Lo. TUBE-LINEMOTOR f3oCIEs a Hiawatha bogie 336 JOURNAL OF THE INST. OF LOCO. ENGINEERS BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 337

base bolster is a major contribution. Within the restrictions caused by the loading gauge and a widely varying passenger load ‘(from 0 to 12 tons for District stock), some finality has perhaps been reached, subject to the development of rubber faspringing and wear avoidance. There remains the maintenance side. The design of the bogie frame has been reviewed from time to time, with a view to eliminating fractures or loose mechanical. connections, and it is thought that, with cross-br?cing and a judicious mixture of riveting and welding, a stable position may have been reached. If wearing surfaces can be eliminated, a more economical basis of overhaul may be achieved.

DEVELOPMENTS * Reference has been made to certain developments in bogie design. The whole question of bogie design is, however, somewhat in a state of flux. Many interesting experiments and suggestions have been made, both in Britain and, even more perhaps, on the Continent and in the United States. Many of these may well be important contributions to the art of bogie design. The measure of their success must be their operating and economic advantages, and these must, on balance, be substantial if departures are to be made from current conventions. Much depends on the duty called for. For main-line railway work with locomotive haulage under British conditions, there is little scope for criticism of the new standard British Railways’ bogie built on strictly conventional lines. This excellent bogie, however, is perhaps not easily adapted to “ motorisation ” or to heavy and frequent braking, nor perhaps is it suitable for speeds, loads, and track conditions current abroad. There is a tendency in the United States-especially for subway services but also for main-line railwaysto develop bogies having the frames inside the wheels (Fig. 21a). This permits a lighter axle owing to reduced bending loads and also permits a more rigid and lighter frame. With this design are often associated axle brakes in disk form mounted between or even outside the wheels. With such bogies, axle-box springing is often omitted and a rubber joint is inserted in two diagonally opposite corners of the frame to secure wheel contact with rail under variable conditions of track level. Although greater unsprung weight results, this disadvantage may be outweighed by the simplicity of the system. There is a general tendency for use of rubber in shear, or combined compression and shear, in place of steel springs, wholly or in part. The London Transport experiment with a rubber-mounted bolster, already described, has been matched by an American development where the vertical loads are taken on helical springs (Fig. 21b). The long-range vertical spiral spring has also been used in the United States to provide easy riding. that is, to provide more “give” for a given impact or load. This device is less suitable with very large variations in passenger load owing to the big vertical travel involved. The possible sideways deflection of a spiral spring to obviate the need for bolster hangers has also been the subject of experiments understood to be successful. 338 JOURNAL OF THE INST. OF LOCO. ENGINEERS

A bogie has been designed by London Transport which uses no steel springs (Fig. 22). The vertical deflections are dealt with by bolster rubber-spring assemblies, similar to those already described but arranged so that the body weight is carried directly on the centres of these assemblies by a new method, instead of on the middle of a bolster. The tractive and braking loads are taken on a cross-beam connecting the spring assembiies. The conventional horn-guide and sprung axle-box is replaced by rubber in shear plus compression, in a form permitting lateral angular movement between axle-box and frame. This bogie has no wearing parts except the wheels.

CONCLUSIONS The foregoing has been a imiew of current and past designs, based mainly upon conventional practices, with a fen references to less conventional variants. Whilst considerable reference has been made to London Transport practice, the problems are common to all bogies in vzrying degrees, and an atfempt has been made to call attention to certain facts in relation to bogies which are peihaps seldom appreciated. Railway mechanical engineers must tend, for well-known and indeed obvious reasons, to avoid experiments. For this reason, bogie design tends to advance slowly, by evolution rather than b revolution. The normal design has proved its practicability over a Yong period, and most efforts have been directed towards improvements in detail rather than to the production of something fundamentally different. As has been indicated, the obvious change from riveting to welding has not been as simple as seemed probable. Arduous conditions have led to efforts to eliminate or reduce wear so that the vehicle can run longer between less expensive overhauls. For electric railways with the motive power embodied in the bogie, and a requirement for heavy and frequent braking, the problem of wear and tear is very important. Simplicity is demanded commensurate with a large proportion of bogies being motored, and bogies cannot be allowed to be as complicated as may be permissible in locomotive practice. Reliability remains, as ever, of first importance in railway work. The problem is to give good iiding for the passengers by practical and economic means at the speeds required, to the runnlng schedules demanded, with the track provided. The last condition is important. Good riding depends on both the mechanical and the civil engineer, and may involve the electrical engineer. The mechanical engineer must not design .regardless of the damage done to the track by the bogies, while the civil engineer can help, say, by welding the rail joints and by good maintenance. If the effects are considered of a bogie which carries the added weight of two electric motors, each weighing 2& tons or more and travelling at, say, 60 m.p.h., turning out of a straight into a curve or following its sinusoidal path along straight track, the mass of such a bogie is large and the lateral forces resulting from its swings about its centre pivot have to be met by the contact of the wheel flange on the side of the rail. How can severe side-cutting of the rail and wear on flanges be other than inevitable? BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 339

How can such side-cutting be other than cumulative in cause and effect? How can vertically-unsprung masses of this rort be other than destructive to track joints and points and crossings? Usually such bogies are inherently bad riding. It would seem desirable to reconsider such designs and to attack the problem as “ bogie plus rails ” rather than to deal only with the bogie. One solution is, after all, simple and even advantageous in some other respects. It is to mount only one motor in a bogie and to use more and, therefore, smaller motors. By the offsetting of the bogie centre thus permitted, some extra weight for improved wheel to rail adhesion may be derived, while the greater number of motors will increase the total proportion of train weight which is carried on motored axles, and will enable a higher rate of acceleration. If the motors can be wholly spring-borne, so much the better. That such an arrangement is less economical than the normal one may be suggested, but does this assessment take account of track wear? There is, too, a real necessity to give the best riding conditions to the passengers whose fares provide the money on which the economics in question are based. A private motor car or a public motor coach does, after all, ride rather well ! Acknowledgments The Author wishes to acknowledge the facilities -available to him through the London Transport Executive for the preparation of this paper, and to thank his many friends in Britain and abroad for information they have made available.

APPENDIX BIBLIOGRAPHY

1. Broadbent, H. R. 1950 Tlir Railway Gazette, vol. 92, p. 69, “ Weight Transfer in Multiple-ITnit Electric Stnck.”

2. Davies, R D. 1938-39 J1. I.C.E., vol. 11,. p. 224, “ Some Experiments on the Lateral Oscillation of Railway Vehicles.” 3. hvirs. R. D., and Cook, A. F. 1948 Proc., I.Mech.E., vol. 158, p. 426, “ Motion of a Railway Axlc..”

4 Spencer, D. W. 1945 J1. Inst. Locomotive Eng., vol. 35, p. 263, “ Notes on Axle Design and Performance.”

DISCUSSION Sir William Stanier, F.R.S. (P.P.) who opened the discussion, said that at the time of his first association with the railways, the rolling stock had four or six wheels; but soon afterwards Dean had introduced the bogie, Fig. 3, with the suspension bolts at the corner of the bogie frame, and had then developed the suspension bolts between the ,wheels. That bogie had given the best riding vehicles that had ever been known at that time; it was so good that a man from the Great Western Railway had made a big reputation in India by introducing the bogie that gave the best riding ever known there. 340 JOURNAL OF THE INST. OF LOCO. ENGINEERS

Fig. 3b did not show the more usual form, which had two suspension bolts at each end on the suspension bar, so that the bogie frame had a suspension bolt on each side of it, which prevented it from twisting. That bogie was of very light construction, and with up to 48 feet stock certainly gave better riding than any other bogie. Unfortunately, traffic conditions had necessitated the building of coaches of 60 feet or over, and the angling of the suspension bolts with the longer vehicles made very uncomfortable riding. Ever since, the railways had been in trouble with bogie design. He knew that one design had pneumatic tyres under the centre. For the main line, it was important to prevent the wheels from developing a double flange; carriage wheels, with more than in. hollow, always led to bad riding. To make two wheels of exactly the same diameter of a cylindrical form was precision work which was not usually obtainable in a wheel shop, and it had been found, after making a number of experiments on the Liverpool and Southport line, when a rapid cinematographic camera had been used to photograph the movement of wheels on the rail, that the cylindrical wheel had the disadvantage that if the flange got against the rail it stopped there until it struck points. The solution was to alter the wheel to a 1 in 100 cone instead of the 1 in 20 cone, which was usual. That gave exceedingly good riding. It had not the same tolerance for wear as the 1 in 20 cone, but it had sufficient tolerance for carriage stock. There was a movement on the railways to build lighter stock. He believed that with lighter stock it would be increasingly necessary to reduce the coning of the wheels, because the sinusoidal movement with a 1 in 20 cone could react seriously on the carriage body. With a light stock the 1 in 100 cone might be a solution. The equaliser body gave a nice riding bogie when it \\as new, but it was very apt, because there was no control on the bogie frame except in the middle, for it to rise up and down, which meant very heavy wear and tear on the parts. He asked whether the Author had found a composition brake block which would maintain its characteristic for braking wheels running in wet weather.

Mr. J. S. Tritton (P.P.) said that the photographs showed twenty-four ways of making a bogie, and the text showed that there were many more. The Author had said that vertical irregularities on the track must be absorbed by the bogie without being transmitted to the body. Ideally, what was wanted in dealing with vertical irregularities (by which was meant rail-joint shocks) was to absorb them where they occurred, i.e., at the point of contact between the tyre and the rail. The best prartical method so far evolved for doing that was the pneumatic tyre, with its swallowing action, but its application to railway practice was limited. There was, however, a compromise-the resilient wheel. Many of those present would have had an opportunity of riding in the American P.C.C. car, streetcars fitted with resilient wheels in which the tyre carried a thin inner steel flange on either side of which was BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 341

mounted a pair of heavy rubber rings. The rubber rings were bonded to the inner flange, and the tyre was held in position by through bolts and dowels on to the wheel centre, but was completely insulated from it. The effectiveness of the resilient wheel in damping the rail-joint shocks was extraordinary-cars could not be heard coming until they were within SO or 10yards. There had, indeed, been complaints that the cars were dangerous on that account, but the complaints were an effective proof of the efficiency of the resilient wheel. Another resilient wheel, developed some twenty-five years previously by the Sentinel company, had been fitted to light rail cars run in the Channel Islands. So far as he knew, that type of wheel had been very effective, and he believed that some of them were still running. That wheel was of a different type, in which the rubber was in compression and not in shear and bonded. He realised that the resilient wheel had severe limitations. The P.C.C. cars had an axle load of 6 tons and a tare weight of 16 tons, but they carried up to sixty seated passengers, and had very easy springing-so much so, that one passenger stepping off the car was sufficient to sway the whole car to an extent noticed by other passengers. So far, the wheels had not been developed beyond that stage, but he thought limitation to a 6-ton axle load could not be contemplated. Sufficient volume of rubber could be incorporated in the resilient type of wheel to give, in due course, an axle load up to railway requirements. In regard to the transverse oscillations and shocks a bogie had to withstand, he agreed with the author's assessment of the amount of the transverse forces; but he had not brought out the fact that if the forces were of a cyclic nature their peaks were usually of momentary duration. He had himself examined many diagrams of flange forces on bogies in the previous few years, and all were characterised by very sharp peaks rising a long way above the normal maximum stresses of the flange forces which occurred. The view was now being taken that the peaks, because they were only of a momentary duration, did not impose nearly such severe. stresses as had at first been thought. Mr. W. A. Agnew (P.P.) said that he must acknowledge paternity for some of the bogies shown in the paper. Some of the earlier types had been inherited from the American engineers who had equipped the District railway and the three London Electric Railway Company's tubes-the Bakerloo, the Piccadilly and the Hampstead lines-in about the year 1905. The cast-steel motor bogies (Fig. 1Sa) and the girdle type of trailer bogie (Fig. 19a), running to easy schedules in sheltered conditions, had served their purpose well. The same motor bogies running on the District line with the equaliser-bar trailer-bogies (Fig. 17a) had given considerable trouble, partly because of the severe conditions of that line, with its long stretches of checked-curves and its numerous crossings, and partly because of the indifferent quality of the materials used. 342 JOLTRNAL OF THE INST. OF LOCO. ENGINEEKS

When the tube lines had been projected into open country, it had become necessary to make radical changes in the equipment to suit the higher speeds. The author had been largely identified with that work and with the introduction of the greatly improved tube trains with all motors and’equipment under floor level. He had also made many improvements in the constructional details of bogies so that they might better withstand the strain thrown on them by the great increase in the volume of traffic and the need to accelerate the services. He had rightly emphasised the advantages obtained by the general adoption of roller-bearings. manganese-steel wearing-surfaces, long bolster hangers, non-metallic brake blocks, and individual brake units. Bogie design was rendered morc difficult on those railways by the restrictions imposed by the narrow load-gauge limits and sharp curves met with in the original tunnel sections, coupled with the need to run at much higher speed on the outlying sections. The Author was pursuing a wise policy in increasing the number of driven axles and thus reducing the hammer blow on the track by lighter motors. An additional advantage was that it permitted thr use of non-metallic brake blocks on such trains, and presented the permanent-way engineer with wheels less destructive to the rails. Maximum-traction bogies with one-motored axle were extensively used on electric tramcars, but there they were running on grooved rails. It was a little difficult to see that a bogie with one axle much more heavily loaded than the other was complet$y satisfactory for all speeds, although it might meet the needs of a particular railway. He asked for particulars of the running of those bogies and of the rubber-spring type. He believed that bolster hangers outside the bogie frame had first been used on certain oversea narrow-gauge railways. Bogie design was in a state of flux throughout the world. Kevision of the whole railway mechanism was receiving much attention, as some modification of the Stephenson railway appeared to be inevitable if railways were to retain their full share of the country’s transport. It was unthinkable that railways, with their inherent safety of working (due to their guided wheels and signal system) and their immrnse capacity for national srrvice, should be allowed to decay. Dr. H. I. Andrews, M.Sc. (Eng.), M.), said that he was very interested in the Author’s plea for consideration of the bogie and the track together, and his consequent recommendation of the use of fully suspended motors, particularly at high speeds. The damage occasioned to the track by a nosc-suspended motor was roughly proportional to the weight of the motor and the square of the speed, and was affected by the length of the wheclbasr; so that for the fast running, which would be probably required in the future, fully suspended motors would appear to be almost inevitable. Before the war, those facts had been considered by the L.M.S. railway, and a more or less experimental design of bogie for very high-speed running had been produced, with fully suspended motors and a simple form of link mechanism included inside the gear wheels. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 343

No difficulty had been found in the incorporation of two 375 h.p. motors, but it meant a heavy bogie, with a total weight of 17 tons. The bogies had now been running for about five years, and while, from the nature of their duties, they had not accumulated great mileage over that period, in all other ways they had been veq. severely tested. Often they had been very considerably o\7er loaded, they had run at speeds up to 85 m.p.h., and had also been used in very fast-rnnning goods trains, where they were subjected to considerable shock; so that they had been very thoroughly tested. Originally there had been one very small defect in erection, and that had 1m.n quickly put right. Later, one of the quill nuts had worked loost., apparently because the type of fastening used was not sufficiently strong, and replacement was necessary. To aimid that difficulty they had been tack-welded in position. and that appeared to have overcome the trouble. Apart from that, no other gear cases hatl been removed, until, after five years, it had been thought desirable to look at them. They had all been taken off, antl inspected, but as nothing could bt found to be done on them, they hatl been put back. He thought that indicated that the only ordinary maintenance needed was to put in oil. It was only fair to say that the flexible drives were fitted with roller bearings; and, just as with axlr-boxes, if a sleeve bearing werc used instead of a roller bearing, it was cheaper, but there was a tendency to greater maintenance cost. The cost of such drivt.3, both in estimate and in .fact, was about ‘Lo per cent of the cost of the corresponding motor. That was a fairly large item, but the’ only alternative, which the Author had mentioned -that of using a large nt1mbr.r of smaller motors spread down the train-also resulted in increased cost and, most certainly, increased maintenance.

Mr. J. 1,. Koffman (A.M.) said that thc Author had said that all studies concerning the horizontal motion of wheels along the track had necessarily been based upon the wheels and track both being in new condition, but an exhaustive and illuminating study of that subject had been published by Heumann (19:37),* who had dealt not only with new \vhrels antl rails but also with wheels and rails at various stages of wear. He commcndrd clow study of the paper to c\~l-yrolling- stock engineer. He entirely agrccd with the Author in regard to rifiid 7~crszi~ loose \\heel sets. In addition, one of the main reasons why loosc wheels wcre not popular was that they did not possess ability to centre themselves in the track. Thus, although they eliminated the sinusoidal motion, they also tended to run crab fashion, with resultant wear in both wheel flanges and rails. As had been pointed nut by Professor Heumann, who had gone into that question in very great detail, the rise of loose wheels for the trailing axle of a bogie would contribute to damping the sinnsoidal motion.

Ilt.urn;it~n, I I. 1937 0rK;tn fiir tliv Fnrtscliritte des Eismlnhnwrsene. vol. 92, p. 149 1943 vol. 98, 11. 221. 344 JOURNAL OF THE INST. OF LOCO. ENGINEERS

He suggested that the Dean bogie was capable of improvement. Its poor riding qualities at high speeds were, to a considerable extent, due to the fact that two widely spaced and rather short hangers were used at each side. By restraining the bogie they tended to transmit its oscillations to the body more than was desirable, whilst running through curves of small radius must 'result in appreciable wear of rails and flanges. If, however, one hanger only were used at each side, then a greater degree of freedom would be achieved and the riding qualities improved. That \had been done in a number of railcars built for oversea railways, incorporating a somewhat similar suspension. The advantages were that the vehicle weight was reduced, while at the same time it was possible to employ a long swing hanger. A further step in the evolution of that type of bogie would k as in Fig. 23, in which there was one long swing hanger at each side which carried the body. The centre pin was pointing downwards and was located to bring it as nearly as possible in line with the axles. The Author had stated that the combined spring system of the bogie should be aperiodic to avoid setting up resonance movements. That applied both to the vertical and to the horizontal motion. In the case of the latter, it might be advisable to use centring means with non-linear characteristics, and thus it would be possible to prevent resonance at the operating speeds. In Fig. 24 the amplitude of the transverse oscillations was plotted as a function of speed for a certain vehicle, and by the use of non-linear centring springs it was possible to move the resonance band into the region of low speeds. In that connection, the question of swing-link length and arrangement should be looked into more fully, for it was most important in bogie design. The use of nests of coil springs mounted outside the bogie frame, as shown in Fig. 2.5, presented an attractive solution, for it was possible to achieve by very simple means good riding qualities in both the vertical and horizontal plane. That solution had originated with Kreissig and had been used by him originally for a railcar built for I.G. Farben. Subsequently, it had been employed for railcars for the Madrid-Saragossa-Alicante railway and later for railcars built in Britain for Kenya and Jamaica. In all cases it had proved extremely satisfactory. The spring calculations were based on simplified assumptions but, more recently, the whole question of horizontal stiffness of coil springs had been dealt with in great detail by Harihgx (1950)* and Borgeaud (1950),t so that accurate calculation was now possible (See page 336 for Fig. 23, 24 and 25-Ed.) The original Reuleaux formula for axles could still be used, but the magnitude of the various safety factors or factors of ignorance which had to be used seemed rather large. Extensive tests carried 'out in Germany had helped to provide a somewhat clearer picture of the stress distribution in actual service. The stresses now allowed as the result of those tests varied according to the type of vehicle and speed, and some of the data were published in Britain.

* Haringx, J. A. 1950 " On Highly Compressible Helical Springs and Rubber Rods and Their Application for Vibration-free Mountings " (Phillips Electrical Company, Eindhoven). t Borgeaud, G. 1950 Festschrift Mirko Nos, Solothurn. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 345

The French pneumatic-tyred bogie was interesting and promised well for city, rapid transit. Now that most countries were rather poor, it would haye the advantage of permitting the use of elevated railways and, in making those silent, eliminate the main objection against that mode of transportation. The number of bogies illustrated conveyed a clear picture of the state of the art. It appeared that there were very few well-established principles, but apart from those, the design was treated as an art rather than a science. In principle, he agreed with Sir Edmund Whittaker that: “ Probably such a state of thing is really quite healthy: first get on with the discoveries in any way possible and let the logic be cleaned up afterwards.” There had been bogies for about 100 yeass, and the time had arrived when the logic should be “cleaned up”-very thoroughly; in that the paper was a valuable contribution. The fundamental scientific work carried out on the subject of vehicles running on rails during the previous fifteen years, particularly that by Heumann (1950, 1951)* and Borgeaud (1944)J was of the utmost importance, and future generations of railway engineers would benefit from it very considerably. He asked whether the Author would enlarge on the subject of one motor per bogie, particularly with regard to any evidence to confirm that such a step was desirable, especially so far as riding qualities were concerned. It had been shown by Heumann (1940)$ that with a bogie having unequally loaded axles, a damping effect was obtained with the heavier loaded axle running at the rear, whereas, if it were running first the bogie would tend to run crab fashion. Mr. A. F. cmk, M.A. said that the statement that clearance must be provided between the axle-box and the bogie frame could be questioned in the light not only of technical evidence but also in the evidence afforded by some of the illustrations in the paper of some recent continental bogie designs. Although it could not be claimed that there was a clear pattern apparent in modern trends in bogie design, it was characteristic of all those designs that a serious effort had been made to reduce or remove altogether the fore-and-aft clearance between the axle-box and the bogie frqe. There was, he believed, an even more recent bogie in use in Germany in which a greater attempt had been made to prevent fore- and-aft movement between the axle-box and the frame than was possible even in the bogies illustrated. Considerable emphasis should be laid’on that. If coned wheels had to be used (and the reasons for using them were very strong), then sinusoidal oscillations arose. The violence of the oscillations could be reduced only by reduction of the coning angle of the treads or reduction of the freedom of yaw-freedom

* Heumann. H. 1950 Elektrische Bahnen, Munich, vol. 21, pp. 81, 112, 159, 182, 255, 282. 1951 vol. 23, pp. 109. 133, 277, 319. 1951 Glasers Annalrn, Berlin, vol. 75, 1). 211. Borgeaud, G. 1944 Schweizerische Technische Zeitschrift, Zurich, Nos. 42-45, p. I. # Heumann, H. 1940 Organ, Berlin, vol. 95, No. 3, p. 43; No. 4 p. 61. 346 JOURNAL OF THE INST. OF LOCO. ENGINEERS to undergo rotations about a vertical axiswhich the axle must have to perform. The main disadvantage in using the 1 in 100 cone treads was that the treads wore very rapidly and that after a comparatively short mileage any advantage derived from the 1 in 100 angle had been lost. A much more attractive proposition was to reduce the freedom of yaw of the axle. That had been apparent for some years in Britain in tramcar bogies-the P.C.C. type which, in addition to its extensive use in the United States, had been for several years running in Blackpool and was to be tried in Leeds-and also in an earlier type of tramcar bogie in which the ordinary axle-box, moving in horns, had been eliminated. The latter was a very desirable attribute, not merely because it prevented the relative movement of the axle and the frame, but also because it eliminated one of the most important points of wear even in a bogie which was fitted with manganese-steel liners. In experiments he was making on an experimental model, the effect of reducing the clearance between the axle-box and the bogie frame was spectacular. From a condition in which violent lateral oscillations (I' hunting ") occurred at a comparatively low speed, the removal of the clearance in the axle-box produced an improvement in which there was practically nb oscillation, even at speeds two or three times that at which a hunting had previously been established which would have made the ride for passengers very uncomfortable. Improvements in the clearance would enable the bogies to be run for a longer mileage before they received maintenance, or, alternatively, if maintenance at approximately the same intervals of time was preferred, to remain in a better condition, and the ride would not deteriorate so much during the normal life of the bogie. Although a reduction in. clearance would necessarily increase the cost of the bogie, the potential reduction in wear might justify the extra expenditure, particularly if, for other reasons, the conventional British type of bogie, which had an axlc-box moving in horns, must be retained.

Mr. G. C. Jackson (A.M.) said'that he was associated with outside contractors building all types of rolling stock. They were not in the fortunate position of being able to carry out trials before going into production, and were often expected to go into production on new designs and to produce them correctly without running trials. That was particularly so where the job was for a line of different gauge from that of the British railways. The Author had been of great help to them on many occasions. Having been engaged on light-weight Diesel traction since well befog the 1939-45 war, he had experienced bogie troubles, particularly with the frames and mostly with all-welded bogies. In changing from riveted bogies to all-welded bogies, many engineers had been in error in taking the riveted bogie as a basis for the all-welded type, and, in the early days, there had been a lack of appreciation of the sudden changes in section which could occur on a welded bogie through that. Far less trouble would have occurred it they had taken the cast-steel bogie as a basis for fabrication. Those two bogies seemed to have BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 347 an essential feature in common : the necessity to avoid sudden changes in section. His, company had had some rather costly experiences with that type of bogie, and were now concentrating on all-welded bogies of box-frame construction. They were attempting to take all tractive, braking, and spring forces centrally on the centre-line of the solebars with no overhanging forces. The bogie was supported entirely on coil springs, and the damping was carried out by hydraulic shock absorbers which were also mounted centrally in the bogie solebar. He asked for opinions on the question of hydraulic shock-absorbers for damping both vertical and side movement of the bogie. They appeared to have the right characteristics, particularly for the side movement, because they had a comparatively small resistance to the initial movement, but a big resistance to the return movement. He agreed with Mr. Tritton’s remarks about resilient wheels. One difficulty with them was in connection with brakes. If a wheel could be designed to take normal axle loads, there would. be so mush rubber in it that the use of ordinary block braking might be impossible because of the high temperature generated in the tyres and its deleterious effects on the rubber inserts. The condition would be aggravated by the use of non-metallic blocks which had not the same thermal conductivity as cast-iron blocks, and therefore tended to produce higher temperatures during braking. Under those conditions, the use of disk or drum brakes or, for electric stock, dynamic braking, would be necessary. As regards the bogie frame the deletion of conventional block braking would be advantageous, as the ideal arrangement of a straight, shallow box-section solebar did not readily lend itself to the usual anchorage points for block-brake rigging, or to the London Transport type of unit cylinder and block, where the anchorage was at approximately axle level. Mr. F. H. Beasant, B.Sc. (A.M.) said that the simplicity and robustness of the nose-suspended traction motor had been stressed, which feature rightly justified its adoption in many cases, in spite of its obvious mechanical shortcomings. The roller-bearing suspension unit shown in Fig. 15 had undoubtedly brought about a great improvement, largely eliminating noise, reducing maintenance costs, and increasing gear life, and he asked whether the only deterrent to its more universal adoption on electric railways had been the difficulties which arose, with running rail-return systems, in ensuring that all current was returned through the axles to the rails without passing through the roller suspension bearings or axle bearings, with consequent damage. The desirability of carrying on springs the whole of the motor weight and, indeed, as much as possible of the gear transmission weight had been stressed by the Author, who had also emphasised the desirability of reducing the mass of electric motors in a bogie to improve riding characteristics. Both of those desirable features could be achieved by the use of cardan-shaft-drive motors, with the development of which he had been closely associated in the previous few yean. By 348 JOURNAL OF THE INST. OF LOCO. ENGINEERS abandonment of the spur gear principle and adoption of a right-angle drive, not only had the motor been carried on springs and a much sounder form of gear unit introduced, but the practical limit of gear ratio had been increased from about 611 to about 81 1. That enabled the motor size, weight and cost to be reduced where the maximum vehicle speed was not required to be more than about 55 m.p.h. The motor was extremely simple in its construction, and the cost of four such motors with an aggregate power on a 1 hour rating of 220 h.p., complete with hypoid gearboxes, was approximately the same as the coSt of two conventional axle-hung motors with gear units and roller suspension units for similar duty. The total weight of the four cardan-shaft-drive motors with gear units would be 5,800 lb., as against the total weight of two orthodox axle-hung motors with gears and suspension sleeves of 7,oO0 lb. and, the total unsprung weight per axle was only 500 lb. as against 2,000 lb. with the conventi,onal motor. Such an arrangement with a motor on each axle made possible higher rates of acceleration an economical provision of electric braking if required. A group of British companies was about to commence the construction of a-substantial number of electric subway cars the bogies of which would follow, to a very large extent, the latest design of bogie evolved by the Author for use on the London Transport system, and would .include many of the very desirable features described in the paper, but would have two cardan-shaft-drive motors, each driving the axle through a propeller shaft and hypoid gear unit, to exploit the advantages described. Several speakers had asked whether it was possible to build a resilient wheel which would withstand the axle loading associated with the railway loads. It was possible, and wheels capable of an axle loading of 12 tons had, in fact, been built. Whether its adoption was correct was purely a question of economics.

Mr. N. Thmeley (A.M.) said that the Author had indicated that the Dean suspension was to be used on the Metropolitan-Vickers gas-turbine locomotive for British railways. That was very definitely not so. While the Metropolitan-Vickers gas-turbine locomotive naturally had a swing-link suspension, the links had no springs at either end and did not directly connect the bogie frame to the body pedestal supports, there being an equaliser beam interposed which was free to pivot in all directions. Also, all link movement was taken by means of bonded rubber universal joints at each end of the links. He was surprised to hear that the Dean-suspension coach-bogies, which had more than fifty years’ use on the Great Western Railway, And many years on the “Cheltenham Flyer,” had “less than optimum riding” above the modest speed of 50 m.p.h. He agreed with the Author that axle-box clearance was most important for god riding. An axle-box had been developed by Metropolitan-Vickers in which the lateral clearances were kept constant by means of rubber-bushed links joining the axle-box to the frame of the locombtive. With that arrangement the thrust faces on the axle-boxes were eliminated and maintenance was greatly reduced. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 349

Mr. R. A. Riddles, C.B.E., (P.P.), observed that the Author’s survey of bogie design indicated that any generally accepted alternative to the conventional bogie was a long way off. The paper emphasised, with particular reference to electric traction, that whatever form a perfect bogie might eventually take, engineers and the travelling public alike had been convinced for a long time that better riding was overdue. He had travelled on the London-Rrighton line in a Pullman car in which the riding was very good, and on looking to see what type of bogie was fitted, he had found that it was the British Railways standard bogie, which had been fitted under a Pullman car to see how it would work. In view of the large amount of experimental bogie design recorded In different parts of the world, little of which so far pointed to conclusive results, it was obvious that any improvement could be only a long-term matter, and since both vehicle and track were equally concerned, there could be no short-cut by using a bogie, designed to run under different track conditions, on British railways. The work was therefore natnrally divided into two parts-short-term and long- term policies. The obvious solution to the short-term policy was to discovei the best bogie available, and that had been done very simply by running comparative trials with all existing bogies and with such instrument- ;ition a? was available, and then, by taking the best out of each, producing a bogie that gave the best possible results with existing knowledge. That had occupied a long time, and had consisted in trying out the bogies under standardised conditions, with records, with both new and well-worn tyres. The results of the tests was the present British Railways’ standard bogie. He was glad that the Author approved of it. It gave a reasonable ride, although they were not finally satisfied with it, and such improvement as hat1 been obtained had entailed increased weight and cost. The long-term policy was therefore more important, but it would be some time before all the an~werswere known. Recognising that it was not only a matter of the bogie itself, a committee had been formed to study the interaction between the track and the vehicle, with representatives from the mechanical and civil engineering departments, under the chairmanship of the director of recearch. The first objective of the committee was to decide how to measure scientifically that elusive quality “ riding,” and to break it down into its different elements. The long-established method in use was not sufficiently analytical for the purpose. There would then be selected a limited number of variations from the conventional designs, and prototypes would be built which could be fully tested against the existing designs, with such improved measuring technique as might be developed. Research of a mole fundamental nature was proceeding, and the universities were assisting the British Railways research department with laboratory work on small-scale models. Motor bogies would be considered equally with trailer bogies; the worn condition was of more importance than the new, and the 350 JOURNAL OF THE INST. OF LOCO. ENGINEERS mileage which could be run before riding conditions became uncomfortable was of the greatest economic importance. In spite of the most.comprehensive work which had been carried out in the United States, France, and elsewhere, there was no short-cut which could be an alternative to dealing with the problems under British conditions. The factor at present missing, and which was absolutely essential, was the means of accurate measurement; that subject was of most absorbing interest and had great possibilities. Much had been said about the flexible'wheel. Only that week he had turned down a project of a fully designed and developed flexible wheel because the cost was practically 50 per cent of the total cost of the coach itself. After hearing the discussion,. he was inclined, in spite of that, to have a set of bogies fitted with such wheels purely for development and experimentation, from which something was certain to be learnt. He was completey satisfied, as were the technical officers of the manufacturers, that a flexible wheel was possible; the design was ready to go into manufacture.

Mr. G. E. Lucas asked what mileage or performance was obtained from the leaf and coil springs fitted to the conventional design of bogie, and whether the rubber form of suspension would give as long, or perhaps longer, life. His experience, which was somewhat limited, was that the rubber units in that type of application generally gave a life much less than that given by mechanical units,

Mr. T. Henry Turner, M.Sc., (M.) said that the Author had omitted two conditions of operation which should be mentioned, in view of the statement that the public motor coach did " ride rather well." Surely there was no public motor coach in which a passenger could write at 60 m.p.h., as was done regularly in the ordinary main- line coaches of a train. The Author had rightly said that the problem must be considered in regard to the bogie plus the rail. The road vehicle never had to go backwards for the same distance and at the same speed, and that was one of the features which had to be considered in the design of the bogie. The " toe-ing in " or castoring action possible with some other types of vehicle could not be considered in bogie design for rail vehicles. A train feature that applied to the electrical bogies for two-thirds of the systems in Britain was that thev must'conduct electricity. Four-rail systems were relatively few in Britain; so that the current was likely to flow through the components of the bogie. Uneveness over rail joints and lateral track irregularities could both be reduced by butt-welding the rails. At the time he had joined the railway service, civil engineers had been afraid of lateral deforma- tion of the track and catastrophic deformation of the track in hot weather. It had been definitely proved that they were dangers about which there need be no worry, where long welded rails were used. He knew of no case where the long welded rail had been catastrophically deformed. The only rails that had so deformed, in railway experience, were those in which there were expansion joints. From French work which had been done on the subject, it would be seen that use of BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 351 the expansion joint was courting catastrophic deformation under thermal expansion in the hotter parts of the year. Hence, there was no reason why rail joint bumps should be considered inevitable. The Author had spoken of axles having lasted longer than they should have done according to theory. Probably that was due to the absence of corrosion. Experience had shown that the average main- line axle would fail by corrosion fatigue in a relatively few years, if it weFe machined. Experience equally showed that thorough painting would preserve it for very many years. Corrosion could shorten the service life to a quarter of what it might have been. Before very expensive pneumatic tyres were adopted, with their greatly increased friction, he hoped that Mr. Tritton’s recommendation would bear fruit, and that the rubber-spring wheel-centre would be considered. With that there was little friction and little unsprung weight. Four or five years previously, when he had approached the biggest rubber undertaking in Britain, they had seen no reason why the success which had been achieved in the tramcars in the United States, Switzerland, and Sweden should not be matched in larger- scale wheels, or why the trouble with heat from braking should not be overcome. The lateral stability of the bogie deserved further experiment. Vertical rigidity was obviously necessary, but by design of the sides so that the one would contract and the other expand (which was possible), he was certain that the winding up which took place in the frame at the expense of abrasion of the rail could be avoided. If roller bearings were used, however, provision would have to be made in the shops to ensure that electric currents did not pass, because it was clear that, in certain of the electrified lines, arcing had been occurring from the race to the rollers.

WRITTEN COMMUNICATIONS Mr. L. Calisch, wrote that the Author had referred to the American P.C.C. tramcar bogie, but he had not mentioned the striking fact that in that bogie the motor driving shaft was at right- angles to the bogie axle, necessitating the use of bevel gears instead of spur gears. The bevel gears used in the P.C.C. bogie were of the hypoid type, similar to those used in rear axles of modern motor cars. The pinion was offset 13 in. below the crown wheel axis and gave extremely quiet running. American P.C.C. tramcars were now being made under licence in Sweden and Belgium. Mr. E. H. Croft wrote that the paper was a very important con- tribution towards the design of rail vehicles. It had given the reasons why bogie stock was essential for passenger coaches; he himself would emphasise that the design of modem Diesel-electric and electric locomotives also embodied bogies, which had taken the place of the dd fixed frames, thus, he would like to think that the paper was the forerunner of many other papers on bogie design. The design 352 JOURNAL OF THE INST. OF LOCO. ENGINEERS of the bogie was of the greatest importance, since upon its design depended the cost of track maintenance, the ease and safety of increasing speeds, and the satisfactory operation of the modern electric and Diesel-electric locomotives. Bogie design had been greatly neglected in Britain and, unfortunately, most of the work and scientific investigation had been left to other countries, Switzerland having played the major part. He would like to think that somewhere in Britain scientific experiments were being carried out on a large scale on bogie design, but he knew of relatively little such work. He was pleased that the Author had definitely referred to the fundamental sinusoidal oscillation of a bogie, which was fundamental to the present bi-conal wheel unit. That was the starting point of much of the trouble. He had held the opinion for many years that the only way to overcome the fundamental tendency to oscillation was by the introduction of a simplified form of differential, and he had for many years been told that it was practically impossible. While realising all the diffitulties he did not agree, and was confident that the ultimate solution would lie in that direction. The papei quite naturally dealt primarily with giving the passenger a quiet, smooth ride but when the bogie principle was applied to the locomotive the riding, while being of importance, was not of such great importance as the reduction of side blows caused by bogie oscillation or curve entry. There was frequent evidence of the enormous severity of transverse blows. While they would not be entirely eliminated until the springing and flexibility of the wheel units in a transverse direction to the motion of the locomotive was fullv considered, they could be greatly reduced by the abolition of bad oscillations, together with reduction of unsprung parts on the axles and/or bogie frames. He had for a great many years advocated the full spring-borne motor, but he realised the difficulty of introducing that. So long a. the use of differentials .was avoided and conal wheels fitted, oscillation could not be avoided and, conseqiiently, the problem was that of damping the oscillations. There was only one fundamental way of damping oscillation-by friction or some form of energy absorption. He favoured suitably designed fore-and-aft bearers as well as transverse bearers, as they tended to introduce a certain friction which, if suitably designed, could be of use in damping nut oscillations. The Author hxl made reference to weight transfer, and he himself would like to emphasise the importance of that in bogie designs for locomotives. IJnless great care and sometimes additional equipment were introduced the modern type of bogie locomotive might easily have 10 per cent, or even more, reduction in effective adhesion for the same total adhesive weight. A further problem for the locomotive which did not arise with the coach was that of the inter-control of the bogies. It was generally realised that there must be control on curves and not on tangent tracks. He did not know whether the Author had considered the question of bogies for locomotives, but he would be interested if he had any comments to make. Mr. W. E. Geh, M.Sc. (Eng.) wrote that flangeway clearances were fixed by the Rerne Conference of 1886 at 0.59 in. minimum and 1.38 in. maximum. The standard clearances in the unworn BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 35:$

condition adopted by the Indian Kailways for lines of 5 ft. 6 in. gauge was approximately 0.62 in. Experiments made in India during 1941 with clearances reduced to 0.31 in. (with appropriate reduction in wheel gauge) showed greatly reduced oscillations of the sprung mass in all modes, and reduced lateral loading of the track. Fig. 1 appeared to show flangeway clearances of 0.44 in. on straight track in the unworn condition. That was appreciably smaller than those in use on the lines of the Railway Executive. The Indian experiments demonstrated that gauge widening was unnecessary, even on the sharpest curves. It resulted in increased resistance to traction and increased tyre tread and flange wear. One of the results of gauge widening on curves was to permit the hauling vehicle to become more skew in the clearances, and that naturally increased angles of attack and slippage. The flanges of intermediate wheels in rigid wheelbases were thinned sufficiently to avoid binding. A definite amount of rotatory control of bogies by springs or by gravity was found to be beneficial in minimising the peaks of the lateral ’ impacts delivered to the track by the vehicle. Some such control.was provided by the Dean bogie shown in Fig. 3, and it was encouraging to see that the bogie in Fig. 22, which also incorporated that feature, was being investigated by the London Traiisport Executive. Mr., R. Howard, wrote that strcss analysis of engineering components, including axles, usually involved (1) experimental determination of the physical properties of the material, and (2) the development (possibly based on experimental methods and ohservctl results) of formulae to give an accurate representation of siresscs set up under operating conditions. In connection with (1) he would be interested to know whether the endurance limit of the axle material used on the London Transport stock had been experimentally determined; and with (2) the form of the corrected Keuleaux formula which enabled accurate determination to be made of axle sizes for stock of lower centre of gravity. From the paper it would appear that, as sometimes happened in engineering practice; observed results did not agree with results predicted by the Keuleaux formula, and the formula ~7ascorrected to hridge the difference . The observed results were perhaps more important than thc predicted ones, and he would, therefore, be interested to know what was the actual life of the axles nsed on London Transport tube stock and how closely it was predicted by the corrected formula. An important consideration in operating electric stock was that of maintenance, and to that end axle-boxes with roller-bearings IIYTC, commonly used. Reference was, however, made to the need for adjustment of the thrust pad in the design of axle-boxes with roller bearings (Fig. 13). Maintenance could be eliminated by the use of roller bearings designed specifically to handle both thrust and radial loads. In one form of such roller bearings, tapered rollers were nsed which had full-line contact between the rollers and races. The area of contact in that type of bearing was greater than was possible with 354 JOURNAL OF THE INST. OF LOCO. ENGINEERS the parallel roller shown in Fig. 13, and resulted in a longer life. That had been shown by the performance of tapered roller bearings with which some London Transport stock was fitted as long ago as 1930, and which were still in service. He was interested in the reference to cross-cornering of axle-boxes in horn guides, because it implied that where relatively weak bogie frame structures were employed the structure could hinge on the roller bearings. In those conditions it was not a question of the wheel sets initiating frame distortion, but of the need for provision of excessive fore-and-aft axle-box clearance to allow for the effects of frame distortion on the wheel sets. There were a number of motor bogies in operation fitted with rigid roller-bearing axle-boxes, in which the fore-and-aft clearance had been almost eliminated. The improved riding resulting from that elimination of axle yaw was well-known. It was also well-known that although the conditions of frame distortion could be partially met by the use of one large spherical roller bearing mounted in the axle-box housing, the use of two spherical bearings in the same housing resulted in a rigid roller-bearjng assembly. Although no reference was made in the paper to the effect of leakage currents on roller bearings, experience had shown that the large contact areas existing between tapered rollers and races were of importance in reducing troubles which might result from it. Mr. G. Nickds wrote that the usual practice seemed to be to incline hangers as s$own in Fig. %a, although there were instances where they were disposed as in Fig. 26a. In the leading bogies of some locomotives the hangers stood vertical as in Fig. 26b. He asked

L a

b FIG.26. DISPOSITIONii:OF HANGERS BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 355

SOLEBAR

FIG.27. PLACINGOF BOLSTERSIDE-CONTROL SPRINGS whether the inclination of the hangers had any effect on the rolling or riding qualities of the bogie. Many bogies were provided with bolster side-control springs placed as in Fig. 27. He asked for the Author’s opinion regarding the Sheffield Twinberrow bogies, in which all dead loads were taken at the side bearings, leaving the centre pin merely a pivot and taking no static load whatever. There had been many failures of motor bogies at comer (solebar and headstock) connections. It had been suggested that they might be largely due to the mounting of the axle gears at one side only of the motor, thus providing a diagonal strain on the bogie frame.

Mr. C. F. J. Ryan, M.B.E., R.A., wrote that he would be interested to know whether the corrected formula used for calculating axle diameters, referred to in the paper, was the same as that which appeared in the paper by Spencer (1945). The principal difference between that formula and that of Reuleaux appeared to lie in the value given to the lateral forces acting on the vehicle. The total value of the forces was assumed not to exceed the value of the force necessary to overturn the vehicle, the value of the axle load being increased by an arbitrary 26 per cent to cover vertical oscillation. In Reuleaux’s formula the height M the centre of gravity was taken as 72 in. and the track gauge as 59 in.; that gave a lateral force of 0.4 times the increased axle load. In the modified formula the actual values of the height of centre of gravity and track gauge were substituted. A basic assumption was that the lateral forces were equally divided between the axles of the vehicle, which was probably far from being true. Both the Reuleaux formula itself and the modified versions were somewhat empirical, and it would be useful to have a more accurate formula, if one were available, based on practical results.

Mr. E. Woodbridge (M.) wrote that the statement I‘ The bolster hangers are usually inclined to the vertical, so that the bolster, with its associated parts, tends to return to a central position after deflection ” was a fallacy. Inclined hangers set up oscillations and kept the coach body bouncing, causing excessive side movement. Straight hangers of a correct length would give better riding, with less side movement and a quicker, smoother return to a central position. 356 JOITRNAL OF THE INST. OF LOCO. ENGINEERS The aim should be to take vertical movement on the springs, and lateral movement on the swing-links, with the wheels hugging the track as centrally as possible. That meant that springing must be flexible and lateral movement free. The oscillation set up by inclined swing-links was largely responsible for the sinusoidal, or hunting, movement of the bogie, which was greatly reduced by fitting vertical links. Some years previously he had fitted a scriber to a bogie bolster with the point marking a plate fixed to the transom. With inclined links the lateral movement was 2; in. When shorter vertical links were fitted the total side movement was reduced to 3 in., the wear of journal bearings was reduced by 75 per cent (in other words, four times longer life), the coach rode better, with less hunting and, consequently, less flange wear and rail side cutting. He considered that springs should be as flexible as space would permit; and the unit deflection of axle-box groups and bolster groups should be the same. That gave easy riding, and the difference in sprung weight carried on the bolster and axle-box springs was sufficient to ensure that there was no periodicity of the springs as a whole. He noted that it wgs the practice with London Transport to line all wearing parts with manganese steel. He had found one wearing part of manganese steel and its mate of cast iron to bg even better, as the two different materials in contact produced a bright, mirror- like surface on each other with practically no wear. In regard to cracks in axles, it must not be overlooked that axles were subject to deflection under load, however small, which produced reversals of stress, causing fatigue. They should therefore be designed so that, as far as possible, they deflected uniformly. He had to deal with repeated broken axles 'on both trailer and motor coaches. The trailer coach axles broke in the centre, and he had found that they were too flexible between wheel seats, and needed to be redesigned to give more uniform deflection. The motor coach axles broke near the wheel seat or gear wheel. The gear wheels were tight up against the wheel centres, causing a leverage on the axle with each reversal of stress. The gear-wheel keyways had then been all rounded and -,$ in. clearance left between gear-wheel hub and wheel centre, which had eliminated the trouble. He had always thought that motor coaches would ride mucli better if they had two, gear wheels on each driven axle, one each side, instead of one. He asked whether that had been tried. Whilst he agreed that good track contributed to good riding, a well designed bogie should ride well on bad track. Too often the designer took for granted that the track would be good instead of designing the bogie so that it would ride well on any track. REPLY TO DISCUSSION .Mr. A. W. Manser wrote, in reply to the discussion ant1 communications,* that Sir William Stader's summary of the rise and * The Council deeply regret that the (lccease of Mr. IV. S. Graff-Baker occurrecl on 15th February 1952. Mr. A. W. Manser, who presented the paper on behalf of Mr. Graff-Baker at the CTeneral Meeting on 4th Januarv 1952. kindly consented to reply to the discussion. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 357 decline of the use of the Dean suspension was interesting, and the suggestion of a reduction in coning of the wheels was being made the subject of some experiments on London Transport stock, and was also a subject on which the research section of the International Union of Railways was taking action. A solution to the problem of composition blocks, for which the coefficient of friction did not vary markedly from the wet to the dry condition, had first been found by the “ Underground ” group in 1931 and much.further experiment and progress had been made since. At the present stage of development, however, there appeared to be advantage in adopting a non-metallic brake block only if certain conditions, which arose on the London Transport Executive’s system, were present. Resilient wheels appeared practicable for tramway axle loads, but experiment with designs for loads double that mentioned by Mr. Tritton had shown the arrangement to be very costly and almost completely ineffective in eliminating noise. The suggestion that lateral forces of very short duration did not impose corresponding stresses seemed illogical and, in the cage of vertical forces, it was the forces of high value and very short duration-i.e., the “ impact ” forces-which he himself considered to be responsible for most axle fatigue failures. In regard to the riding of the asymmetric bogies introduced on both the tube and surface line stocks of the London Transport Executive in 1937-38, the riding of the tube stock bogies was very good-in fact, it was considered the best riding of any of the London Transport Executive’s tube stock bogies. The same could not be said of the surface stock bogie which, although it did not exhibit vicious tendencies, did not ride as well as some other types. The reason for the difference was difficult to define and it would be wrong in the present state of knowledge to attribute all the responsibility for the difference to the fact that, for design reasons, the tube stock bogie carried the motor on the inner axle while the surface stock bogie had the outer axle motored. Concerning rubber suspension developments, experience over a fair period had been obtained only with a rubber bolster suspension. It had been put into service in 1947 and had now run some 225,000 miles. The riding was still good, and no doubt the arrangement was worth adoption as it eliminated many wearing parts. The rubber axle-suspension was a recent development and, although its initial performance was most impressive, further experience was necessary before it could be more widely recommended. Reduction of the unsprung weight borne by the motored axles was no doubt desirable, but became progressively less important with the reduction in size of the motors which followed the increasv in proportion of motored axles. It was apparent that, by the time the position was reacbed when all axles were motored, the effective unsprung mass of the traction motor might represent less than one-third of the weight of the wheel set, so that the additional complication of the quill drive would become of doubtful value. The damage to track by nose- suspended motors was probably exaggerated. The nose-suspended motor was a convenient scapegoat, and it must not be forgotten that 358 JOURNAL OF THE INST. OF LOCO. ENGINEERS the motor did not commence to leap about without provocation. To argue from so limited an application and experience as the drive units used on the dynamometer car was against all his experience. The review of suspension development made by Mr. Koffman might point the way to improvements in riding of high-speed vehicles, but in all such matters the maintenance cost must be remembered and simplification and the elimination of wearing parts must be the aim, rather than improvement of technical performance at the cost of disproportionate complications. It was in that respect that the rubber suspension seemed so promising.

The question of justification of the I‘ one motor per bogie ” policy was purely a compromise between what was desirable from an engineering standpoint, what was necessary to meet operating requirements, and what was permissible in capital cost and subsequent maintenance. For high rates of acceleration, the motive power had to be spread over a number of axles and, by gutting one motor in each bogie, a design was achieved which was less distorted by space consjder2tions ihan one with two motors on the same bogie, and was less damaging to the permanent way: It was not thought justifiable to generalise about the riding qualities of bogies with unequally loaded axles; too many variables existed to enable any certain conclusion to be drawn. The London Transport tube stock of that type was an exceptionally good riding bogie on reasonable track although it had the motored axle in the inner position, so that the leading axle of each vehicle was the more lightly loaded one and, according to conclusions quoted by Mr. Koffman, might be expected to proceed crabwise. The reduction in fore-and-aft clearance of the axle-boxes in the bogie frame had been advocated. For carriage stock which had to run on track where the incidence of curves was high, there was no doubt that attempts to reduce the axle-box clearances were damaging from many aspects, and experience to that end on the London Transport system had been similar to that obtained in earlier years on the Great Western system. The question of the use of wheels having a reduced coning had been mentioned; experiments on that subject were difficult to conduct without the introduction of disturbing factors such as the running of some stock with standard coning over track with 1 in 100 inclination, thus invalidating the conclusions which might be drawn as to wear of the track and/or its effect on the stock with 1 in 100 coned wheels. He would not be prepared to accept deductions from the behaviour of models; there were bound to be too many departures from the conditions required for dynamic similarity. He agreed that additional initial cost might, be justifiable in a bogie which was subsequently going to cost less in maintenance; the rubber-sprung bogie shown in Fig. 22 had been proposed on that basis. In regard to the error made in taking the riveted bogie as a pattern for the bogie of welded construction, in the light of experience he agreed that it would have been better to have taken the cast-steel bogie as a pattern. His opinion concerning the use of hydraulic shock absorbers was that they would be most costly to maintain and unsatisfactory in performance; that seemed a logical conclusion from experience with BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 359 heavy road vehicles. Mr. Jackson had mentioned the difficulty in mounting unit brake cylinders without introducing undesirable stresses in the solebar. That. consideration, together with the very heavy cost of the unit system, indicated the likelihood of successful development with disk or drum brakes-combined with dynamic braking in the case ’bf electric stock-being well within the bounds of possibility. Mr. Beasant had made a good case for the small frame-mounted motor with its possibilities of higher rates of acceleration (with all axles motored) and all that he had said was factual, but he himself wished to point out that increased rates of acceleration-although necessitating no increase (in fact, possibly a decrease) in total energy consumption-did increase the maximum intensity of loading on sub- stations and, therefore, might require the provision of additional power supply equipment, not only because of the increase in maximum demand, but also by reason of the increased root-mean-square value of the current taken. He was grateful to Mr. Thorneley for directing attention to a misunderstanding concerning the suspension used on the Metropolitan- Vickers gas-turbine locomotive. The axle-box development referred to by Mr. Thornelcy was interesting and he would be glad to have particulars of experience with the arrangement after a lengthy period of service. MY. Riddles had mentioned the good riding of a British Kailways standard bogie. on the Brighton line of the Southern Region and he agreed that the bogie had admirable riding qualities. He was glad that attention had been directed to the necessity for giving priority of consideration to the worn condition when considering bogie riding and to the impracticability of deducing answers on the performance of stock from experience with different stock on different track. If British Railways proceeded with the resilient wheel experiment, the results would be of general interest and it might be that with a different design more success might be achieved than in the London Transport experiment.

Mr. Turner had stipulated that bogies for) electric stock “ must conduct electricity ” but, surely, it was up to the designer to provide a proper path for the current and not leave it to find its way through any component (including roller bearings) which might seem convenient. He agreed that thc use of long welded rails would remove one of the main causes which initiated irregular riding. Although the failure of axles by corrosion fatigue might have a considerable influence under some conditions, his opinion was that by far the greatest number of axles withdrawn from service because of fatigue cracks had had cracks induced by shock stresses which were incapable of calculation and could be allowed for in the Reuleaux or any other formula only by increasing the “ factor of ignorance.” Perhaps, with the development of strain gauge technique and the possibility of transmitting strain gauge results from a rotating axle, it would become feasible to measure the stresses in an axle in service and determine the values of shock stresses when the bogie was passing through cross-overs and round curves with brakes heavily applied. He 360 JOURNAL OF THE INST. OF LOCO. ENGINEERS felt that the results of such investigations would probably provide the explanation for the high incidence of fatigue fractures in axles operating under arduous service conditions. The fact that on the P.C.C. bogie the motors were frame-mounted and drove through hypoid gear units was agreed to be one of importance. The necessity for research in bogie design had been urged, but it ,appeared from Mr. Riddles’s remarks that British Railways were proceeding with the matter in a systematic way, and development was also taking place on London Transport. He could not agree with Mr. Croft on the likelihood of a form of differential providing improved riding of wheel sets. The numerous complications-some of them potentially dangerous-which a differential would introduce made the consideration of such a complication unwarrantable. He agreed on the desirability of reducing unsprung weight and of the advisability of using some form of frictional damping for radial oscillation of the bogie rather than introducing the complication of hydraulic dampers. Bogies for use under Diesel and electric locomotives presented different design problems from those for coaching stock and there it was also a big field for development. The six-wheeled bogies in use under British Railways’ locomotive I‘ 10,000 ” and “ 10,001 ” give very good riding and, because of that, the justification for some of the more complex designs where part of the weight was wasted on pony axles seemed difficult to appreciate. The information concerning reduced flangeway clearances which had been tried out in India was interesting. Gauge widening up to in. on curves of ten chains radius and under, was the standard practice on the London Transport Executive’s system and was considered necessary for optimum riding and minimum flange and rail wear. The London Transport Executive’s gauge was, however, kept Q in. below the nominal 4 ft. 84 in. on the straight. In regard to the design of axles, the experimental determination of the physical properties of the steel was a routine measure in the manufacture of axles for London Transport, but fatigue tests were not undertaken as there was considered to be little value in such tests except at full scale. Concerning the design of roller-bearing axle-boxes, he agreed that boxes containing bearings of the tapered roller type had been used by London Transport since 1930. The mileage run with them was good but not exceptional; and the design had two disadvantages (1) the additional time and labour involved in removing the boxes for wheel- turning operations every 50,000 miles or so, and (2) the exclusion of the use of ultrasonic testing equipment because of the dispersive effect of the screw-threaded holes in the end of the axleby which holes the bearing locking plate was secured. The suggestion that tapered roller bearings were capable of carrying leakage currents without damage .was not borne out by his experience. Evidence had been obtained that the passage of currents of a small order-less than 10 amps.-could be damaging to quite large bearings if permitted to continue. The proper course was to provide. a path for the current and to provide, at some point, an insulating barrier in the path via the bearing. BOGIE DESIGN WITH REFERENCE TO ELECTRIC RAILWAYS 361

In his experience the inclination of the bolster hangers did affect the riding qualities of the bogie; though the effect of a change in inclination was not invariably the same, different designs of bogies required different angles of inclination for optimum riding. For coaching stock, the outward inclination of the bolster hangers (as from top to bottom) was of advantage in counteracting body roll on entering curves. He considered the use of side control springs was undesirable, in that it was difficult to make such a spring-control system aperiodic. He believed that the failure of bogie headstock-solebar connections was due to racking of the bogie frame, and unconnected with an question as to where the axle gears were mounted; indeed, as, wit g an axle-hung motor, the traction motor and axle formed a closed unit for all forces except torque, it was apparent that the position of the gear could have no effect outside the motor-axle combination. Mr. Ryan was correct in supposing that the corrected Reuleaux formula was the one referred to by Spencer (1948). A more accurate formula was unlikely to be contrived until strain-gauge measurements on axles in service had been taken and the value of what he himself believed to be the critical unknown-the shock stresses-had been determined. Mr. Woodbridge's opinion about bolster hangers was too much of a generalisation; there was not one answer for all bogies in the matter of bolster hanger inclination. In about 1934 a bogie had been constructed, at the direction of Mr. W. A. Agnew, on which provision had been made for adjustment of the bolster hangers for experiment. Tests made with various loads and with the hangers inclined at various angles either side of the vertical had resulted in the inclination of 1 in 6 outward from top to bottom being chosen for bogies of that design. He agreed that the combination of manganese steel and cast iron was good for hard wear but in many cases it was impracticable to arrange a simple replacement part in cast iron and the manganese-manganese combination was, therefore, commonly used. As to the possibility of an improvement in riding resulting from the use of dual gears, he could not see that such an arrangement would have any effect on the riding of the bogie and he was unaware of such an arrangement having been tried with that object in view.