1950.] 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



11- 



RAIL^VAY CARRIAGE AND WAGON SPRINGS. 



{With Engravings. Plate IV.) 



Oil Railway Carriage and Wagon Springs. By -Mr. J. AV". Adams. 

 — (Paper read iit the Institution of Mechanical Engineers.) 



The object of this paper is to discuss and analyse the various 

 fiirnis and descriptions of springs now in use in Railway Carriages 

 and Wagons; pointing out, to the best of the writer's knowledge 

 and experience, their advantages and defects, and suggesting such 

 imjirovements in the details" as will lead to better effect and eco- 

 nomy in their use and manufacture. 



Huffing and bearing springs are applied to carriages and wagons 

 in order to al)sorb and neutralise as far as possible the force and 

 moiuentumof tlie shocks to wliich tlie vehicles are exposed in tlieir 

 ordinary work. A perfect hearing or buffing s))ring would be that 

 wliich would absorl) the entire power aiul space of the blow with- 

 out disturbing the inertia of the vehicle. This in practice is 

 w b(»lly inipossilile, from the varying loads on bearing springs and 

 varying force on buffing springs. In bearing springs tlie nearest 

 approach to perfection is in the modern first-class carriage, wlieie 

 the disproportion of total weight between loaded and unloaded is 

 less than in any otiier vehicle. 



At the present time, as far as the writer is aware, there is no 

 rule or fm-mula by wliich engineers or manufacturers can ascertain 

 the true form, weiglit, or cpiality of material to be used for effec- 

 tually springing a railway veliicle, and conrecpiently the goods and 

 luiner.il traffic iif tlie country, averaging fn. in 35 to 40 cwt. per 

 spring, is now carried on springs which vary in weight from 35 to 

 110 lb. each. 



The primary object being in all cases to discriminate between 

 good and bad material, the writer has endeavoured to test the 

 relative ipiality of spring steel converted from Swedish and from 

 Kii:;lish iron. For this purpose bars of ordinary spring steel were 

 iirocured from various makers, some being English and the others 

 Swedish ; the bars were all 3 inches wide and /^-inch thick. These 

 bars were cut to e(|u.al lengths, marked, and then made into springs 

 and tem|>ered in the ordinary manner; each of the sjirings con- 

 sisting of a single iilate turned over into an eye at each end, and 

 18 inches long between the centres of the eyes. These springs 

 were then proved in the jiresence of Mr. W. P. Marshall, by 

 means of pressure applied at the centre of each spring, the spring 

 being sup|iorted by a pin passed through the eye at each end, 

 which rested on rollers to allow the ends to be drawn together 

 freely when the spring deflected. The results were as follows — 



No. Weigtit. 



■ 1. 



\'* cwt. 

 •-'U „ 



English. 



Deflection. 

 .. 1 inch . 



B'Okell 



Permaiunt 



Se'. 

 , , g incit 

 .. 24 inch 



From the foregoing experiments it ajipears that the elasticity 

 sustaining power, and toughness of the Knglish steel was much 

 greater than that manufactured from the Swedish iron. 



The I.aniiiiitted Spring is the most common form for the springs 

 of railway vehicles, ((insisting of a number of plates, the taper 

 being given by reducing the plates successively in length. 



The ])riuciple for regulating the taper of the spring is to obtain 

 an e(]ual amount of strain or deflection from each paiticle of ma- 

 terial. If some parts of the spring are deflected less than others, 

 the amount of m,;terial might lie reduced in those parts witiiuut 

 impairing the sustaiiuiig power of the sjiring. 



A laminated s|;ring may be tajiered either in breadth or thick- 

 ness, but if parallel in thi.-kness and all the |)lates the same length, 

 each (date should be uniforniiy tapered in breadth, so that ea('h 

 half of every plate would be a triangle. In jiractice the plates of 

 laminated s|iriiigs are made parallel in breadtli and thickness, inas- 

 much as the jiarallel bar is the most economical form, and the 



taper is obtained, as before expressed, by the different lengths of 

 plates. 



If a spring consisted of only one plate, parallel in breadth but 

 tapered in thickness, such taper should be in the form of a para- 

 bola, as the strength is in proportion to the square of the thick- 

 ness. This form is shown in fig. 2, Plate IV, by the part AA. 



Fig. 1 represents one-half of an ordinary wagon bearing spring. 

 Fig. 2 is the same spring pressed flat, but supposing the plates not 

 to elide over one another. 



If the spring consisted of a number of very thin parallel plates, 

 the correct form would be a uniform taper in thickness from the 

 centre towards the ends, as shown by the jiortion BB in fig. 2, 

 because the strength of each part of the spring would depend upon 

 the number of plates at that part. In practice the most correct 

 form of spring is between the two forms of the triangle and the 

 parabola, but is nearer the triangle, as the thickness of the plates 

 bears only a small proportion to the average length. 



The spring shown in fig. 1 is 3 ft. 3 in. long, 3 in. wide, and 

 4'fi} inches thick in the centre, and consists of 15 plates j^-inch 

 thick, excepting only the outside plates, which are |-inch, according 

 to the usual ]iractice, to allow for the plate not being supported by 

 plates on both sides. 



If this Sjiring were a single plate of the same total strength it 

 would be only 14 inch thick at the centre, and in the form of the 

 parabola -A.A in fig. 2; but as it consists of a number of plates, the 

 outline must be a line lieyoml that curve. 



The straight line BB in fig. 2 is drawn outside the curve, giving 

 a uniform taper from tlie centre of the spring to the end of the 

 second plate, leaving the top plate its full thickness to the end. 

 This line BB appears suitable to be adapted for the practical out- 

 line of the spring, as the deviation from correctness is only very 

 small and gives a slight diminution in strength at the (juarter 



I length D, which is advisable in practice, because tlie centre C is 

 usually weakened bya^-inch rivet hole, reducing the strength one- 

 eighth at that point. 



j The line BIJ is transferred from fig. 2 to the curved spring in 



' fig. 1 by dividing the length of the top plate into Ki e<iual parts by 

 the lines from 1 to lii, which are drawn vertical in fig. 2, and 



I radiating to the centre of the curve of the spring in fig. 1. These 

 lines being made of e(|ual length in both cases give the curved line 

 BB in fig. 1. The end of the top plate is lengthened and turned 



j down at E to give a bearing to the spring. 



I The writer has in practice set out all springs reipiired by him, 

 by drawing tlirougli the extreme points C and E a circular arc of 

 the same radius as the top plate of the spring. The line obtained 

 by this method is a singular instance of how near practice has 

 approached theory by this simjile method, the extreme difference 

 being only J-inch. 



•Tlie line 1111 is obtained in the same manner as before described, 

 excepting that the spring is not tapered to the centre, but to a set- 

 off of 2 inches from the centri>, viz., from C to H. This is the form 

 universally adopted, but it is clearly incorrect, as the centre is 

 made proportionately weaker than the remainder of the spring, as 

 well as being further weakened by the rivet hole thi'ough the 

 centre. 



The true and correct form of spring would be, that the centre of 

 the spring should be at H, and the plates connected not by a rivet 

 but with a narrow hoop. In practice the spring is clipped to and 

 bears on the axle-box at H, and conse(iuently the mass of steel II 

 to C is entirely wasted. 



In two plates of steel of the same length and breadth but of 

 different thickness, the amount of deflection caused by the same 

 weights is in proportion to the cube of the thickness, although the 

 breaking strength is in proportion to the square of the thickness; 

 consequently if one spring were made with plates double the thick- 

 ness of those of another s)iring, the first would require only 

 one-eighth the number of plates, viz., one-eighth the weight of 

 material to support the load with the same amount of deflection ; 

 but in that case the extent of the displacement of the particles of 

 the steel in the thick plates would be double of that in the thin 

 plate.s, and in the practical application of thick plates to springs it 

 is nei^essary to limit the deflection within the above extent, as the 

 double amount of deflection would break or strain the particles, 

 presuming that in the thin plates the particles were being strained 

 to a reasonable e.vtent. 



The Wagnn Bearing Spring in ordinary use on the Midlam', 

 London and N(nth Western, and other railways is shown in fig. 1, 

 and is 3 ft. 3 in. long, 64 in. camber, 4^ii in. thick, and 3 in. wide, 

 consisting of 15 plates of which 2 are 3-inch and th rest -jji-inch 

 thick, and the spring averages in weight about 93 lb. 



17 



