1845.J 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL 



379 



RAILWAY RETAINING WALLS. 

 [The following extracts are taken from a valuable paper by Mr. Demp- 

 s«y, just published amonK the papers ou subjects connected with the 

 duties of the corps of royal enginters. Tlie author has honoured us by 

 frequent quotations from the C'ivil Kogineer and Architect's Journal, 

 which, of course, are here omitted.] 



The only cases in which artificial retaining walls appear desirable are 

 those in which it is actually or nearly impossible to interfere with the 

 surface, which would otherwise be required, lu order to substitute cut- 

 ting or embankment at the natural slope of the material of which they 

 may be composed. Many instances are recorded of the failure of these 

 structures, which has commonly resulted from the saturation and conse- 

 quent swelling of the earth behind them ; and these effects have occurred 

 frequently despite the most judiciously selected forms and materials, and 

 the best attainable system of back drainage. Indeed, unless the material 

 be adapted to stand by itself, be thoroughly impervious to water, or so 

 completely drained that very little reaches the back of the wall, it is cer- 

 tain that this uncontrollable agent will make its way through the work, 

 and produce sooner or later the disastrous conse(|ueuces which have 

 already marred the designs of railway engineers. 



As applicable to cuttings, artificial retaining walls, unless they can 

 be constructed under the most favourable circumstances, are best secured 

 by arches thrown between them, or by other intermediate resistance-, 

 vrhich are required to act as abutments between the two walls, and pre- 

 Tent their forward movement towards each other. With this addition, 

 it is evident that the structure becomes a bridge, or nearly assimilates to 

 one, and is palpably inapplicable to very long walls, except at a tremend- 

 ous cost. 



As applied to the feet of embankments, where the material is of that 

 cohesive but slippery nature that a simultaneous movement of the entire 

 mass may be apprehended, low and strong retaining walls are useful, 

 and present frequently a judicious expedient, though in many cases less 

 advisable than piling or wattling. As protections against the sea, re- 

 taining walls to cover the lower or entire face of the embankment, the sta- 

 bility of the work ; but in such and other cases, if any thing more than a 

 mere breakwater or rude collection of heavy stones, the wall, sloping back 

 to the embankment, 'becomes rather a facing of masonry than an inde- 

 pendent construction, and contributes to the stability of the work only by 

 its artificial cohesiveness and greater weight. Applied to the height of 

 embankments formed to a steeper slope than that at which they would 

 stand independently, retaining walls may be considered never advisable, 

 and would be infinitely better abandoned for a cheap construction of 

 viaduct. 



M'heo their adoption, however, becomes imperative, the stability of 

 retaining walls will depend upon the nature of the soil behind them, and 

 the means taken for its drainage ; and upon the form of wall adopted, and 

 the manner and materials of its construction. 



The amount and kind of pressure of the earth against the wall which 

 retains it are evidently affected by the angle of repose, or natural slope 

 of the earth; also by the quantity of moisture it will imbibe, the propor- 

 tion of this which it will retain, the extent to which its absorption of water 

 will cause expansion of the mass, and by other circumstances. 



The natural slopes of some kinds of earth have been observed expe- 

 rimentally, and are thus recorded : 



1 . Fine dry sand 35° 30' 



2. Gravel (dry ?) 37° 



3. Loose shingle, perfectly dry 39° 



4. Common earth, pulverized and dry 46° 50' 



5. ditto, slightly damp 54° 



G. Earih, the most dense and compact 55° 



Of these results. No. 1 is the mean of experiments recorded by Ron- 

 delet. Barlow, and Hope; No. 2 is on the authority of Lieut. Hope; 

 No. 3 is recorded by General Pasley ; Nos. 4 and 5 by Rondelet;aud 

 No. by Barlow. 



According to the specific gravity of these substances, it appears that 

 the weight of the triangular section (one foot thick), which is bounded 

 by the vertical back of a wall 10 feet high, a horizontal line level with 

 the top of it, and the natural slope af the material, will be about as 

 follows : 



1 . Sand 6800 lb. 



2. G ravel 6;i 50 



3. Shingle fcGOO 



4. Earth, dry, 4.^00 



5. Ditto, damp 3700 



6. Ditto, dense 35S0 



The mere weight of earth to be sustained thus appears to vary very 

 widely according to its constitution and state of dryness or moisture; 

 but this comparison of weights, forming only one element of the calcula- 

 tion, does not furnish any estimate of the actual resistance which the 

 wall is required to exert. This will evidently be reduced by the cohe- 

 sion of the moving mass, and by the friction between this mass and the 

 natural slope of that portion which would remain stationary in the absence 

 of the wall; but, on the other hand, it is increased iu a great degree by 

 the action of water within not only this moving mass, but also the other- 

 wise quiescent mass beneath it. 



In a slate of perfect dryness, and disregarding the withholding elfert 

 of cohesion and friction, the maximum power required in the wall would 

 be represented by the actual weight of the retained earth, supposing this 

 weight to act against the vertical plane of the wall with the same force 

 that it would exert upon a horizontal plane that supported it. And the 

 power thus required in the wall might be immediately calculated for each 

 section of its altitude; but the moment that water is introduced wilbia 

 the retained material, a multitude of other considerations arise, which no 

 theory has yet furnished the means of estiinaling, and which require a 

 series of experiments to enable us to predelerniiuc with any chance of 

 accuracy. 



Upon the methods of draining retaining walls, reference may be made 

 to the account of the Blisworth cutting, given in tlie second section of this 

 paper ; also to the description of the mode of repairing the walls of the 

 London and Birmingham railway, contained in the seventh volume of the 

 Professional Papers of the Corps of Royal Engineers. 



Among the various considerations to be entertained in the desigouDg 

 of retaining walls, that of the influence of scaton should not be disre- 

 garded. A wall built during a dry season, or after a long drought, will 

 incur an augmented pressure when the earth becomes saturated with 

 water ; and on the other hand, if built during a wet season, and backed 

 up with wet earth, it will be subject to a shrinking away of this backing 

 when subsequently drained. The engineer can provide against injury 

 from these circumstances only by making the wall as far as possible self- 

 supporting, so that any retiring of the earth behind shall not endanger its 

 stability ; and at the same time, oll'ering every facility for the water to 

 find its way through the wall, and for discharging it thence into the foun- 

 dation drains. The most perfect wall would be that which should be 

 throughout its whole surface completely permeable by water, so that no 

 accumulation of that fluid could occur behind it, and which should yet 

 effectually retain the particles of earth. Upon such a wall the ellect of 

 the earth would be reduced into that of its mere weight, and experiments 

 would be needed only to determine the best arrangement of bricks and 

 mortar, or other materials, for resisting this action. 



As to the best and most economical forms for these walls, we are 

 enabled to describe some which have been constructed, and have fulfilled 

 the purpose of their construction, and we may also refer to the objections 

 against some forms which have been suggested and adopted ; but on this 

 part of the subject experiments are also much wanted, and under this 

 conviction, all must regret the premature loss of Lieut. Hope, whose 

 skilful labours in this department promised so many valuable results to 

 civil as to military engineering. 



Of plane walls, five different forms have been constructed : first, 

 having vertical faces; second, having one vertical and one inclined face, 

 converging towards the top, and presenting either of these surfaces to 

 the retained earth ; third, having both faces inclined, and converging 

 towards the top ; fourth, having ime vertical face, the other inclined and 

 converging towards the bottom ; fifth, having its faces inclined and 

 parallel. 



Each of these forms is sometimes varied, by curving the inclined 

 lines : then, the second form will present a concave surface to the re- 

 tained material, or otherwise a concave exterior surface ; the third will 

 have a concave surface ; the fourth form will have a convex surface 

 against the earth, and a plane vertical surface externalU ; and the fifth 

 form will present a convex surface to the earth, and a concentric concave 

 surface externally. 



()l these forms, the first three are evidently adapted to stand alone ; 

 while the fourth and fifth will depend for their stability partly upon the 

 outward thrust which the retained material will exert against them. 



It is frequently found that the middle or upper part of a retaining 

 wall fails first, and will be thrust some inches forward, while the lower 

 part of the wall remains firm. This occurrence, which, considering the 

 greater weight opposed to the lower part of the wall, cannot be attributed 

 to the mere weight of the earth, is probably owing partly to the superior 

 strata being less dense, and therefore more rapidly saturated with water 

 than the lower strata. This would lead to the necessity of proving es- 

 pecially for the complete surface drainage of the retained district. 



In other instances of defect, the whole wall is moved bodily forward, 

 and sometimes with very little fracture. This was the case with a por- 

 tion of the wall built on the line of the Birmingham, Bristol, and Thames 

 .lunction Railway, which, for a length of 40 or 50 yards, was pushed 

 forward off the foundation, to a distance of 8 or 10 feet, the wall still 

 standing. The failure was pronounced by Mr. Vignolles, who examined 

 the works at the time, to have resulted from the accumulation of water, 

 which, " having no outlet, had settled the earth against the back of the 

 retaining w:ill, turning the clay into mud, and by the great additionai 

 weight forcing it into the position in which it then appeared." 



There is no doubt that the ultimate stability of retaining walls is 

 affected by the state of dryness of the masonry when the earth is filled 

 in behind it; and also by the manner in which the filling in is conducted. 



Reverting to the forms for retaining walls, it may be remarked that 

 a preference has been shown towards the tifth, with some little modifi- 

 cation. By railway engineers tliis form is nsually reduced in thickness 

 at the top, by steps on the inner face. The experiments made at Dublin, 

 uniler the direction of Sir J. Burgoyne and the Board of Public Works 

 of Ireliiud, are quoted by Mr. Vignules as fully proving that the " parallel 

 batlcnng-waU'' is the one which uUers llie most support, with the least 



