(Puper N . 3 7 . o 05)

‘‘ The Nachinery used in the Manufacture of Cordite.”
T H E Committee appointed by herMajesty’s Government to produce a smokeless explosive for military and naval purposes, consisting of Sir Frederick Abel, Bart., K.C.B., Hon. M. Inst. C.E., Chairman, Professor James Dewar, and Dr. August DuprB, withCaptain Thomson as Secretary, having practically settled the composition as well as the shape in which it was to be used, requested Sir William Anderson, K.C.B., Vice-President Inst. C.E., to assist thembydesigningsuitable machinery for its commercial production. H e accepted the invitation, and, assisted by the Author, had made some progress inthematter,whenin 1889 hewas appointed to his present office as Director General of the Ordnance Factories. The Author then took up the subject alone, and has since kept in close touch with it. He now proposes to place on record a description of the machinery a t present in general use; and, so far as they are interesting or instructive, the steps by which the various designs or processes were arrived at. Whilst , many ideas and suggestions were due to Sir William Anderson and the Author, a considerable portion were likewise due to the ExplosivesCommittee and their staff in the earlier stages, and later to the officers and staff i n charge of themanufacturing establishment atWaltham Abbey. The names of Dr. Kellner, Chemist to the War Office, Major Nathan,and Mr. C. Frewen Jenkin, Assoc. M. Inst. C.E., may be specially mentioncd in this connection. Cordite, as now made, consists of 58 parts of nitro-glycerine, 37 parts of pn-cotton, 5 parts of mineral jelly, and 20-83 parts of acetone, which is merelyasolvent and is driven off in the drying process, so forming no part of the ultimate product. Thesematerialsare made by processes which it is outside the province of this Paper to describe. The form in which the explosive is for the most part produced, namely, that of solid of various diameters,was determinedbythe cordsorstrings

Delivered by ICEVirtualLibrary.com to: IP: On: Tue, 13 Sep 2011 11:38:49





[JIinutes of

Explosives Committee as the most suitable, the sizes being arrangedtosuitthevarious weapons for whichthe explosive was to be used. Thus, for example, for the 0.303-inch LeeMetford rifle the nominal diameter of the cords is 0 0375 inch, and for the12-inchnaval gun it is0.5 inch. The cords are cut into pieces of suitable length and are bundled together in sufficient number to form the requisite charge for the cartridge. The nominal sizes generally made are 0.03 inch,0.0375inch, 0.05 inch, 0.075 inch, 0.1 inch, 0 . 1 5 inch, 0.2 inch, 0.3 inch, 0.4 inch, 0 . 5 inch indiameter;buttheyarenotusually described by these figures, but by thoseobtained by multiplying them by 100. Thus 0.0375 diameter inch is called 32, and 0 . 3 inch diameter is called 30. The length in inches is usually given also, as the denominator of a fraction of which the numerator is the size. Thus signifies cordite 0.2 inch in diameter cut into pieces nominally 14 inches long. The uniformity of diameter is an important factor in the production of uniform ballistics, upon Government which good shooting so largelydepends;andthe specification providesfor only a small permissible variation i n this respect, but somewhat more latitude is given in regard to the length.For use i n blankcartridges it is found thatthe best form of the explosive is that of thin circular wafers, made by cutting round cord into transverse slices. The process described in the Paper, under the sevensections into which it may be divided, is that carried on at the Government factories of Waltham Abbey and Woolwich Arsenal ; it is probable that private manufacturers conduct some of their operations ina slightly different manner.


This process is always performed by hand. The gun-cotton is produced in the factory in theform of finely divided pulp, which is compressed with slight pressure into cylinders about G inches long and 3 inchesin diameter, and in this state is dried; the production of dust, which reason for so compressing i t being that the forms a dangerous element in the drying process, as it so easily takes fire, is very much lessened. Formerly the dry gun-cotton was removed to a store; now the stoves are allowed to cool, and are used as stores, so as to avoid both the danger and labour attached to moving. Afterdrying,thecylindersare weighed outinto brass-lined wooden boxes, provided withsuitable covers, which are then conveyed to the nitro-glycerine factory, where the proper
Delivered by ICEVirtualLibrary.com to: IP: On: Tue, 13 Sep 2011 11:38:49




amount of nitro-glycerine is poured into them, and incorporated by hand till a proper mixture or solution has taken place. The boxes are then covered and are removed to the paste store to wait the next stage of the process. Until a few years ago it was the custom to store a certain quantity of nitro-glycerine in a special building set apart for the purpose, the object having been to give time for the complete separation of the washing water, and there to mix it with the gun-cotton; but since the explosion in this building i n May, 1894, no storage is permitted; and the mixing is performed in the building where the last stage the manufacture of of the nitro-glycerine is carried on, by which means the only free nitro-glycerineabout is that in actual prcjcess of manufacture. The danger of explosion is greatly diminished by this arrangeexplosivesubstances ment, for it is found that mixing the two results i n a comparatively inert compound, which, though highly inflammable, is not easy to explode, and may be safely transported from place to place. I f therefore any storage is required, i t takes the form of this mixture, instead of the much more sensitive and dangerous nitro-glycerine and gun-cotton. This fact also enables the factory to be divided if required, keeping the nitro-glycerine section completely by itself. The gun-cotton may then be sent to the nitro-glycerine section, and be returned as f‘ paste,” as it is usually termed. This method was adopted bythe Government when the nitro-glycerine factory was wrecked in the explosion referred to, and the ‘6 paste ” was made of Government gun-cotton sent in the wet state to private nitro-glycerine factories, whence the mixture was returned to worked into cordite a t Waltham. be

This operation is performed by aid the of machines, the so function of which is tothoroughlymixalltheingredients as to yield a homogeneous mass of uniform quality. The system universally used, so far as the Author is aware, is that of Messrs. Werner, Pfleiderer and Perkins,whose machines, originally designed for mixing and kneading various materials, have proved, withslight modification, to be exactlysuited to this purpose. The Author is indebted to Mr. Paul Pfleiderer for the following description of the machines. In a trough of double semi-cylindrical shape, Figs. 1, Plate 2, with a prismaticupward extension, twosets of blades revolve interdependently at different speeds about axes corresponding
Delivered by ICEVirtualLibrary.com to: IP: On: Tue, 13 Sep 2011 11:38:49

the trough with its revolving wheels and blades is tilted in its entirety. the other by a crossed belt. or elongated boss of the disk. become rapidly distorted. CJfinutes of with those of the two semi-cylinders. The direction in which the blades revolve. one by an open. and are squeezed and moulded into more or less perfect cylinders.screwed at its end. areheldat amaximum distance by two s d p s or set collars fixed to the shaft. the blades withthe wheels fixed on their axles can readily be lifted off also for periodical cleaning. For ordinary cleaning and for discharging the finished mass after each operation. The materials to be incorporated are poured into the top. as shown by dotted outline. The boss of the handwheel on the front side and the boss of the back pulley on the back. The top being raised. The blades touch and intersect each other’s charge. and facility of emptying and cleaning. Thearrangementis such that. is reversed by a simple friction clutch. the handwheel is revolving in the same direction. and the handwheel forms the nut for this screw. It consists of two loose pulleys with a fixed diskbetweenthem.theattendant engages thepulleywhich revolves i n that sense. but their orbits never intersect. The peculiar shapeandsetting of the blades arethe outcome of many years of experiment.169.130 On: Tue. and by stopping it the attendant stops the machine. or hopper. Boththemixingandthedischarging operations are facilitated by reversing the direction in which the mixers work. By a handwheel the attendant forces the disk into or out of grip with either pulley. Adapting them to a large number of materials to be treated has finally produced the bestshapes for efficiency of working.143. as they emerge from their moulds (the semi-cylinders of whichthe bottom part is composed). Delivered by ICEVirtualLibrary.72 ANDERSON ON THE MANUFACTURE OF CORDITE. A pair of rings connected by pins passing loosely through the disk keep the two pulleys a t a constant distance. over which i t fits loosely. and the blades do not touch-an important feature where the mass is explosive. pushing the one out of gripwiththedisk before the other can possibly be engaged. 13 Sep 2011 11:38:49 . of the trough . Many of the present shapes may resall ships’ propellers. while when the machine is working in either direction. The pulleys are driven.com to: IP: 194. between the faces of which set collars their axial movements are confined. they fall between the revolving blades.by screwing the handwheel in the direction in which he wishes the machines to be driven. but they have all been derived by evolution from an inclined plane with elliptical contours. The disk is secured on theshaft of the machine by a feather. the loose pulley runs on a sleeve. either against or from each other. which.

] ANDERSON ON T H E NAh'UFACTURE OF CORDITE. Accidents. and is taken straigh! to the presses. and the connections with the water-main anddrainare made by flexible pipes. however. the size for the most part employed was that capable of holdingabout 75 Ibs. i n amuch higher degree than is often obtained by more complicated shapes.130 On: Tue. it is essential that the shape of the blade should be i n each case carefully adapted to the nature of it must depend largely on the the mass to be treated.169. As the material under operation itself plays a most important part in bringing about the desired mixing and kneading action. Delivered by ICEVirtualLibrary.andthereby monotony in the mixing movements. as it is important to use i t promptly before the acetone can evaporate. or 2 lbs. so that.143. The cover is put on i n order to prevent evaporation OI the acetone as much as possible. 73 intersected and displaced.com to: IP: 194. A lively interchange of particles sets in. never occur. they effectually prevent the formation of 6' dead centres '' or " cores. when the manufacture was commenced on a large scale. though wise.and oohesiveness of the latter. A drenching-pipe is placed over each machine. which in their paths are rubbed and pressed against and into one%notherinall directions..corporation. I n the process of incorporation the paste is first put into the machine and the acetone is poured on it.'' Their ever-varyingrelativeanglesand positions prevent a frequent of repetition of the same phase or relativestate. The shapes of the blades are so chosen. and on its adhesive qualities in regard to the metal of which the mixer consists. and now a machine to contain 150 lbs. and highly scientific kinematic motions which are liable to get out of order. whichhas been previously wellstrained. The speed of the large sized machine is about 31 revolutions per minute.After 34 hours themineraljelly. Various sizes of these machines have been used. orsixteen charges perworking week. whichare difficult to clean. 13 Sep 2011 11:38:49 . andthe machine is started. whenthe process is continued for 3$ hours more. and the precaution is. and holding only 1 lb. thattogetherwiththeir differential speeds. practically superfluous. but subsequently.PL'OCecdings. should any accident take place. At the close of the operation the explosive is emptied into suitable covered boxes. so thatthe machine may be tiltedeasilywithoutbreakingany joints. Theoriginalexperiments were conducted with machines worked by hand. the charges in all themachines can be quickly drowned. of explosive. so that the whole period of incorporation is 7 hours. has been generally adopted. About three charges can be incorporated by eachmachine in 24 hours. The bottom of the machines is water-jacketed to prevent undue heating duringin. is added .

The remedy is. Accordingly a screw was inserted between the hydraulic cylinder and thepressing chamber with a nut and thrust-bearingthat Delivered by ICEVirtualLibrary. allthe machinery is designed to fulfil this end. and is usually solid. should be uniform. and one of a size suitable for making rifle cordite was projected. though for some purposes. and having n consistency aboutequaltothat of stiff dough. and to place a brush collector near the belt. when all cause for anxiety is removed. and a t a given speed. [JIinutes of A curious source of danger was discovered in the factthat frictional electricity is developed by the belts of these machinos. however.130 On: Tue. namely by forcing it through a die of theproper size and shape . is ready for conversion into the form of cord.74 ANDERSON ON THE NASUFACTURE OF CORDITE. of small dimensions and of steel. and upon theresults of these thedesign was based.. so that it could be easily used i n conjunction with a small hand screw-press. The original laboratory apparatus used in the experiments a t Woolwich was.com to: IP: 194. Later. Experiments were first tried to ascertain the pressure necessary to force the dough through a hole of given size. and consequentlytherate of emission of the cord.also e a r t h d . it was decided to make a hydraulic press . but. likethis. 13 Sep 2011 11:38:49 . Thereappearstobebut one way in which this can be done. requiring larger quantity of the cordite. As the pressure necessary was found to be high.169. such as for primers. it is only necessary to connect the machines electrically to earth. I n this particular case t h e cord is invariably circular in section. it became necessary to make a press of commercial dimensions.and a close-fitting plunger capable of being forced into it. having a die of therequired dimensions at the end.143. the dough will forced out through the be die in the shape of the hole through it. which could deal larger with quantities. so that if the cylinder is iilled with the doughand the plunger ispressed in. The obvious plan is to use a cylinder of suitable size. it was considered that the regulation of the pressing should be effected by positive gearing such as a screw. Soon. in view of the subsequent operation which the cord would have to undergo. and so enable experiments sizes and a greater to be tried on heavy guns. however. accordingly. so that the motion of the plunger. and. and sparks could be taken from them. it takes the form of a tube. The material having been properly incorporated. hydraulic presses were employed. PRESSING. simple.

and. owing to improvements in the details of manufacture. so that the friction of its passage through the die was sufficient to ignite it. whereit is still used. imperfectly dissolved gun-cotton. 13 Sep 2011 11:38:49 .0375 inch i n diameter. pierced with a number of $-inch holes. &C. the pressure required. and it was erected inthe temporary experimental shed a t Woolwich Arsenal. when the Author was present.com to: IP: 194. The axis of the cylinder was horizontal inthisfirst press.169. or the results might have been disastrous.130 On: Tue. by which means any tendency to uneven motion might be regulated and corrected. 75 could be rotated by suitable gearing. and although. owing to the elasticity of the explosive. Kellner. It was found that a considerable quantity of impurity. the emission of the cordwas arrestedby an obstruction in the die.butthat to provide some meansfor preventingsuch excess of pressure. so that the hydraulic cylinder could be dispensed with and the screw alone used-a planwhich answered well. in the apex of which the die was placed. as experimentally obtained.a very muchsmalleramount of Delivered by ICEVirtualLibrary.under it was a shallow conical recess in the cylinder cover. On two occasions. the cord was shot out for an instant at great speed. the pressure in the cylinder was increased to such an extent that it dislodged the obstruction . The pressing cylinder was made 2 inches i n diameter with a stroke of 6 inches. a suggestion of Dr..Proceedings.143. and thereupon. but the press continued working. and the die was 0. to hold about 1 lb. so the pressing cylinders were constructedin such way that they a could be easily removed and fresh gauze disks could be inserted as required. was considerably i n excess of that now proved to be needful .] ANDERSON ON THE YANUPACTURE OF CORDITE. probably owing to recent modification in the composition and plasticity of the explosive. and they required frequent cleaning. This system is still in use. It was soon found that the combined hydraulic cylinder and screw was not a success. who The difficulty wasmetby chamber of the proposed a strainer between thedieandthe cylinder-a plan which proved successful. There were. before it could be stopped.but it fortunately happened that. This press waseventually removed to Waltham. The arrangement consisted of a steel plate & inch thick. This showed that corditewas a very safeexplosiveto manufacture-a factthat it was advisable hasoften been demonstratedsince. was caught by these strainers. fixed in the bottom of the cylinder and covered on thecylindersidewith a disk of iron-wire gauze. Fortunately it didnotstrike back throughthe die on either occasion and fire thechargeinthecylinder. howerer. still difficulties to be overcome. of the explosive.

13 Sep 2011 11:38:49 . but they also interfere with the quality and regularityof the product. A second point was. are placed in aspecial receptacle. be multiple dies were successfullyintroduced. the first point of difference is that it was vertical instead of horizontal-% plan whichproved a decided improvement. [Minutes of impurity reaches the material than formerly. and which was guided i n a cast-iron framework . It may. of course. The plunger was attached to a steel cross-head. The cleaning is performed by soaking in acetone till all the cordite dissolved. and. In this new design the point was raised whether it would not be possible to renderthe process of manufacture continuousinstead of thereciprocatingsysteminvolvedby a press. it was decided to design a larger press more suitable for the quantity of cordite that would be required-a duty which was undertaken by the Author. The renewal of a disk is found to be desirable after about six charges in the smaller-and three charges in the larger presses. and were revolved by suitablegearing. A third point was the old question of screw versus direct hydraulicaction. but as it became necessary to prosecute trials to a greater extent in the latter direction. and are taken away at intervals forcleaning.com to: IP: 194. and which has been invariably followed since. Plate 2. thedifficulties originally found with them having been surmounted. the nut and thrustblock were carried by a bearing in the upper cross-piece of this frame. but the results were not mentioned that eventually satisfactory. Fig. Comparing it with the former one. and the plunger stroke was nominally 12 inches. The result was the machine now usually called the Intermediate press.143. and experiments were made in this direction. which. Experiments had previously been made chiefly with rifles.130 On: Tue. and it was decided toadhere to theoriginal method. The pressingcylinder was made 6 inches i n diameter. and area t present used to a considerable extent. necessarily wasted acertain amountof time in the return stroke and in refilling or charging the pressing cylinder. whether multiple dies could beused with advantage. The disks. no plan was arrived a t t h a t gave any promise of success. the screw again carried the day on account of the uniformity of its action. after use. in view the subsequent of processes.as it must beremembered that foreign substances not only constitutea source of danger. when dry.76 AMDERSON ON THE MANUFACTURE OF CORDITE. however.169. and then brushing is with a scrubbing-brush. and only to a limited extent with large guns . so that the chargeof explosive was equivalent to about lO& lbs. 2. on the upper end of which the screw was forged. I n order to shorten Delivered by ICEVirtualLibrary. After much thought had been bestowed on the subject. there is even now sufficient to render the strainer necessary .

When therefore thereturnstrokewas ended. when the plunger had reached the end of the downward or pressing stroke. leaving a clear space between the plunger and the cylinder. The result was that the upward thrust of the press was taken upon this water. nut. so that a new method was introducedwhichgavethe pressseveral important advantages. this valve was allowed to close. and which was made capable of sliding up and down on the columns between fixed stops at the top and bottom. the whole framework was quickly raised. so that. exerting sufficient pressure to raise thewhole frame.com to: IP: 194. As soon as the return stroke was completed. upon which could thedrivingpinion of thegearing on theslidingframe travel. andquickly was turned on totheupperside forced the charge out of the so-called charging-cylinder into the pressing-cylinder. A hydraulic cylinder was attached to the top of the two main steel columns of the press.and pressure waterwasturned on tothe annulus of the cylinder. The connection with the driving gear was preserved by means of a vertical shaft with a long feather in it.which could be automatically opened by the valve lever when the water from the top of the cylinder was required to be exhausted .to. The pressing-cylinder was necessarily heavy.130 On: Tue. On reversing the hydraulic valve. bodily about 12 inches.143. a quick-return motion was provided. or even shortly before that. into which a light brass cylinder filled with the explosive material was a t once introduced. and a cast-iron ram proceeding downwards from i t was attached to the cast-iron frame already alluded . and its removal forcharging was notvery easy. and remained so until started the by attendant for a fresh pressing stroke. with the screw. the gear was automatically stopped i n a similar manner. 77 as far as possible the time of the cycle of operations. The annulus formed between the ram and the piston of the cylinder was sufficient to allow of the water. the gear was automatically reversed and the quick return proceeded without any intervention of the attendant. The hydraulic slide-valve had a non-return valve in connection with the upper end of the cylinder. and. supplied from the ordinary waterworks mains in thefactory. the water of the cylinder.] ANDERSON ON THE WANUFACTURE OF CORDITE.169. 13 Sep 2011 11:38:49 . by means of open and crossed belts runningon different-sized pulleys on the driving-shaft . the gear was then started and the pressing commenced. so that. by attaching a safety-valve to the upper Delivered by ICEVirtualLibrary.gearingandplunger.Proceedings. when the frame was down ready for pressing to commence. until the framecame down to the bottom stops. and thus the water in the cylinder was imprisoned. and a simple arrangement of belt-striking apparatus was contrived.

provision was made for one speed only for the advance of the screw. thereby assisting in no small degree in securing uniformity the in finished product-a matter of the highest importance.130 On: Tue. it wasarrangedto useadifferent size of cylinder for each size of cord. This therefore constituted a safety device. asmultiple dies have been adopted. The descent of theplunger compresses thisintothe pressing-cylinder. if any undue. I n order to simplify themechanical arrangements. the safety-valve would relieve the plunger by allowing the whole frame to rise. gives a very good indication of the uniformity of the incorporated material and the correctness of its consistency.2 inchdiameter is tobe made. but as the speed of emission of the cord was required to be the same for most ofi. minutes of end of the cylinder loaded to a given pressure.the sizes made. I n addition. for instance. the uniformity of that pressure throughout the charge.78 ANDERSON ON THE MANUFAUTURE OF CORDITE. &C.143. and worked well. the 6-inch cylinder was used forcordite of 0 3 inch diameter. after withdrawalof the plunger by the hydraulic arrangement. 13 Sep 2011 11:38:49 . ifcordite 0. so that. and a 4-inch one for cordite of 0. with the important difference thattheoutput of the press is doubled. Thus.com to: IP: 194.the speed of the press has been increased by an alteration to the gearing. at any moment the actual pressure on the explosive could be seen. It was subsequently sent to Waltham Abbey. so that. a double die of this size is put into the 6-inch cylinder. But the arrangement had yet another advantage. the diameter being calculated to givethe necessary speed of emission. and is provided with a side-feeding hopper . that the charging cylinder is now fixed. and prevented any possibility of an accident such as that recorded i n connection withthe first press. where it is still in regular use.friction.2 inch diameter. and for successive charges under similar conditions. because.and then the screwbegins to workas before.stress was put upon the pressing plunger through such a cause as the obstruction of the die. a treble die of 0. This press was set up at Woolwich. Although the absolute pressure is of no great importance. for by attaching a pressure-gauge to the upper end of the cylinder. without decreasing the speed of emission. a slide at the bottom of the hopper being withdrawn.2 inch diameter. the pressing-cylinder and about 9 inches of the charging-cylinder are at once filled with loose incorporated material. or a double one Delivered by ICEVirtualLibrary. and produces the same effect as using a single die with a 4-inch cylinder.169. The only material alteration that has been made is. Thecylinderis now seldom or never changed. the necessary allowancesbeing made forrelative areas of cylinders.

designed i n such a way that it can be taken in and out easily. containing the gauze and the die. and they have never yet beenadopted i n makingthis size. which. of which many would required. but a plain screw was eventually found to answer better. 13 Sep 2011 11:38:49 . can now be used with a6-inch cylinder. Each press had two cylinders. Regulation of the speed was of even greater importance rifle cordite than for the larger for sizes. of which one was finally adopted. waB again used and produced the same effect.] ANDERSON ON THE MANUFACTURE OF CORDITE. andthe Delivered by ICEVirtualLibrary. be Two or three schemes were prepared by the Author.143. for accelerating this process . 79 of 0 . was a t once decided upon . consisted of a cast-iron frameworkfixed to the floor. is replaced. as in the former press.Proceedings. after changing the gauze. This press. and thus a screw press was almost imperative. of a similar design to that in the larger press. so that while one was i n use in the press. which. and when in can be locked i n position SO as to be kept truly concentric with the plunger. and this at once made i t necessary to prepare a design for the presses for making the rifle cordite. if not all. and although the question of multiple dies was again mooted. Fig. 3. the private factories.130 On: Tue. and no time was lost between successive charges. the and good qualities of the new explosive having been established. The slow motion for pressingandthequickreturn were given by pulleys of different diameters on the worm shaft. The strainers in the bottom of the cylinders are removed for cleaning by unscrewing a largeplug provided with handles. Plate 2. the following method was devised.withslightalterationsand improvementssuggested by subsequent experience. no hydraulic arrangement was required. having boen successfully started. This plug was at first made with an interrupted screw. The press. having a crosspiece in which recess was formed for holding the a pressing-cylinder. is that still in use by the GoTernment. andthus a further increase inoutputhas been obtained. being small. can easily be removed for filling. it was abandoned. which worked so well in the intermediatepress. like a breech-block. which was bored out on itsunder side. The vertical arrangement. forming a short cylinder. 3 inch diameter. Inasmuch the as cylinders.com to: IP: 194. and which. as the safety device and pressure indicator were even more important in thispress than in the other. The Screw was operated by a worm and worm-wheel. and so advanced the screw'. The steel screw terminated in a cross-head. driven by open and crossed belts. Striking gear for the belts. and a t most.169. but. the erection of a permanentfactory on a large scale a t Walthanl Abbey was decided upon. revolved the nut. the other was being filled with the dough.

It consists of a small hydraulic cylinder and plunger.com to: IP: 194. 13 Sep 2011 11:38:49 . so stopping the press. One of thefirstwas a new press larger than any yet made. Mr. the improvements have been altogether i n small details. has now been increased. [Minutes of top of theplunger rod was expanded to form apiston fitting this cylinder. guided by the experience whichwascontinuallybeing gained. and oil was the liquid used to fill it . and therefore. Frewen Jenkin wasappointed mechanical engineer to the establishment. the pressure of the screw was transmitted to the plunger through the medium of that liquid. Thus. whereas oil is objectionable. the press is automatically stopped.143.[toabout 376 feetperminute. by means of a flexible copper pipe. the plunger is forced out. and packed with a U leather. on the pressure rising abnormally. which is found togive less trouble from leakage. With these two types of presses the Waltham Abbey factory commenced to work. by means of which. Formakingupleakagedueto possible blowing off from t h e safety-valve or otherwise. but in the more recent presses the buffer has been increased t o 6 inches in diameter.130 On: Tue. should any leak out and reach the explosive. Since the first presses of this type were made. which wasgoriginally fixed a t 100 feet per minute. a small hand-pump and tank were fixed to the cylinder. it does no harm. with the safety buffer lengthened and modified. because. withthebeltstrikinggear. connected by means of a suitable weighted lever and bell-crank. with the buffer cylinder. No new principle of actionwasinvolved in its arrangement. overcoming the resistance due to the weighted lever. The buffer cylinder was originally made 4 inches in diameter. The speed of emission of the cord. if the pressure in the buffer cylinder (as it may be called) becomes excessive.SO ANDERSON ON THE MANUFACTURE OF CORDITE. and a s nominal plunger-stroke of 6 inches. a pressure-gauge and safety-valve having been attached so as to communicate withtheliquid. Moreover waterhas been substituted for oil. and. C.partlyby increasingthe size of the pressing-cylinder. and moves over the belt f striking bar by means of the bell-crank. The cylinders a t present used have a diameter of 2 inches. he introduced various new designs in connection with the manufacture. and partly by the speed of the gear. Shortly afterwards.169. so as to act as it Delivered by ICEVirtualLibrary. and. and i t may be generally described as a large rifle press. I may be mentioned that the last new design of the intermediate press has also had a similar arrangement added to it. i n order to keep the pressure lower. Anadditionalsafeguardhas been since added to this press.the desired result wasobtained. The cylinder having been filled with liquid.

the whole thus forming a kind of stool on the top of the ram. all accidentsthathave happened inpressingcorditehave been i n connection with direct-acting hydraulic presses. and they have onlyserved to show the comparative safety of the manufacture. when dry. ] IP: 194.169. though. It will be seen thatthis press is muchsimplerthanthe screw press. so that no special pumping plant was required. except those of minor importance recorded in describing the original imperfect machine.130 G On: Tue. 4. The cord is emitted at the bottom of the cylinder as usual. as far as the Author is aware. which was supplied a t a higher pressure thanthatavailableat Woolwich.proceedings. perhaps. illustratesthearrangement as designed a t Waltham and made at Woolwich Arsenal. but it seems reasonable to suppose that they would never happen were screw presses exclusively used. inaddition to its other functions. and capable of taking a pressing-cylinder 8 inches diameter by about 16. and [THE INST. inches stroke. The ram is 22& inches diameter. The Waltham press is of ordinary type.143. Plate 2.] ANDERSON ON THE MANUFACTURE OF CORDITE. and the quantity of explosive dealt with per charge was equivalent to about 21 lbs. It is true that none of the accidents have been serious. and he does not recall anyinstances of such an occurrence with a screwpress. It is nevertheless difficult of explanation. The next step was simplification of a the pressingmachinery for the large sizes of cordite. The . The pressing plunger is fixed to the top crosshead casting.com to: VOL. by departing from gear-driven screws and substituting an ordinary hydraulic press.asthecylinder position in accordance with the motion of the screw. as well as the means of knowing the pressure a fact.E. Delivered by IICEVirtualLibrary.afterthemanner of the lintermediate press. that. C X Y X ~ . 13 Sep 2011 11:38:49 . per square inch. and it would appearthatthesafety precautions introduced into the screwpress are of necessitypresent in the hydraulic press.connections to the hydraulic valve for controlling these motions shifted its werenecessarily of a flexible type. The water for workingthehydrauliccylinders of theintermediate and large presses was obtained at Waltham from the mains of the East London Water Company. upon which the pressing-cylinder is carried in such a way that it canbeeasilytaken inandout of the press. C. Fig. and has projecting from its top four short steel columns surmounted by a cast-iron plate. It would seem that speed had an effect on the liability to accident. and passes away between the short supportingcolumns. constructed to work with the available waterworks pressure of about 135 lbs. 81 hydraulicraising-and-loweringapparatus. one atall times. The pressing-cylinder was in this case 8 inches i n diameter with a nominal plunger-stroke of 12 inches.

so that one is being filled while the other is in the press. heating is set up either by friction or.pug-mills made. and it is therefore somewhat similar to the rifle press i n manner of working.as it happensthatthe mains is not by any means steady. if the press is allowed to work too fast. sufficient heat being generated to ignite the mass locally and cause explosions of varying degrees of violence . the dough was cut through with a wire. the result being that the cords made bythedirect-acting presses areliable tobe of irregular diameter . 13 Sep 2011 11:38:49 . possibly by sudden compression of the air in the interstices of the dough. and the system was abandoned in favour Delivered by ICEVirtualLibrary. When full the cylinder was slightly slacked.169. 83. [Minutes of that.com to: IP: 194. as has been suggested. the rule being that the greater speed the smaller the diameterof the the cord. and which might prove detrimental to the finished cordite. A newdesign of intermediate press has just been made. when no doubt the result willbe satisfactory. so that the effect of a possible explosion may be confined to the machine itself.143. but the arrangement of the parts is more like thatof the original intermediate press. Filling the Cylinders. being forced out as a homogeneous mass into the press cylinder. i n which the cylinders are removable by means of a small crane.-It was originally thought that it would be very important toconsolidate the material wellbefore pressing. in which were the loose incorporated dough was placed. The experience at Waltharn has also demonstrated clearlythe bad effect of irregular motionon the water-pressure in the resulting cordite. The results. and the cylinder was removed to the press. so as to get rid of any air that might lodge in the interstices. These hydraulic presses are arranged so that the cylinders can be taken out andfilled in any convenient place .130 On: Tue. Particulars of output. I n consequence of this stateof things the authorities at Walthant enclose thelargedirect-acting presses i n mantlets. With this object. and hence the damage is never very great.. The screw presses are not protected inany way.82 ANDERSON ON THE MANUFACTURE OF CORDITE. of the different presses are given in the Table on p. whileanother one was attached to the pug-mill inits place. to save time. were not satisfactory. which was attached for the purpose to the delivery nozzle of the mill. because the size is considerably affected by the speed of emission. and. each press has two cylinders. &C. Hence steps are being taken to obtain a perfectly steady source of supply. however. which will necessarily vary if the pressure varies. a curious circumstance being that only a smallpart of thecharge ever explodes. in the early experiments two.andoperate them from the outside.

4 .130 On: Tue... .com to: IP: 194. . .. ..I .S ... . id V I I:: & - - e -0 I '2 Delivered by ICEVirtualLibrary..169...... .143...-.. .. 13 Sep 2011 11:38:49 m 2 -h l4 n c . z . .. * E ......... . .... ..

though the output per strokewas reduced. and is found to answer well. 13 Sep 2011 11:38:49 . Like the direct-acting hydraulic presses. into which the pressing cylinder can be slipped. and the stroke was. and is then withdrawn. by means of which a similar result was obtained. by its means.Themainframehas a holder similar to that in the screw press itself. As soon as this had acted to its full capacity. Finally this was superseded by a small hydraulic rammer. Subsequently.81 ANDERSON ON THE MANUFACTURE OF CORDITE. compressing the dough into i t with moderate force. and it therefore becomes important to regulate them so that they cannot beworked too quickly. It consists of a smallvertical press. The heads of the Delivered by ICEVirtualLibrary.thechargebeingfrequently rammed down whilebeingputinwith a wooden rammer. these rammers have occasionally given rise to slight explosions. and the plunger operated as before. completed. the reason probably being that the pressure is much less thanthat in the small presses. The cylinders of the larger presses are usually made of phosphor bronze . A similar system was employed for charging the cylindersof the intermediate press and the direct-acting hydraulic press . The vacant space in the cylinder is again filled. a lever rammer of wood was devised.130 On: Tue.com to: IP: 194. Minutes of of fillingbyhand. high-pressure water was let in from an accumulator supplied by a small set of pumps. the plunger is brought down. and i t is then removed to the press.143. as in cases where the cylinders are removed for charging. havingan overhead inverted cylinder with a plunger. but itwas eventually abandoned in the case of the former for the method described. mild steel is sometimes used. which is now used in all the large screw presses. probably for the reasons already suggested. I n one set of hydraulic presses designed by the Author for a factory where low-pressure water was available. and does not score. the method adopted was to fill the pressing-cylinder with loose dough after the manner of the large screw presses. working downwards from it. which is still in use. and a special filling press was dispensed with. but finally hardened wrought iron was adopted.for the rifle-press cylinders. the head of which is somewhat smaller than the bore of the pressingcylinder. By this method highpressure water was saved. casewhich were at that time made of mild steel. A third repetition of the process leaves the cylinder almost full. experiMaterials of the C~linders. and to give it a preliminary compression byturning on the low-pressurewater.169. and carries a funnel to assist filling in the dough.-Difficulties were a t first enced through scoring of the cylinders used in the small presses. The cylinder is first filled. though where lightness is required.

i t mayeven requirevariation to suit circumstances in order to secureauniform resultinthe dried product. The cordite for the rifle cartridges did not suffer from this cause. soon evident that this method was only applicable to the smaller sizes . 85 plungers are in all cases made of phosphor bronze. storing. the mixture of gun-cotton and nitro-glycerinewas obtained from other factories. The dies for the smaller sizes are made of hardened steel. The first and obvious method WRS to coil it on reels from which it could be afterwards unwound as required. the cordite. because then the reeling machine could be always made tosynchronize withit. Having been. it was found that a different-sized die was required i n pressing the larger sizes to produce dry cordite of the correct size. has to be collected in a manner suitable for the subsequent processes. and for the larger sizes of phosphor bronze.143. Forinstance. so that suitable flanges automatic conwere provided. and for it the reel system was in every respect the most convenient.Theoriginal reels were 9 inches in diameter by G inches wide. and making into cartridges. a t anyrate. and has been invariably adopted. i t was found that the larger sizes obtained a permanent set to the curve of the reel. however. because it has been found that. 1894. and therefore theoriginal horizontal machine press had a reeling apparatus attached to it. as the drying was performed when on the reels.com to: IP: 194.169. The reeling machine had an trivance for continually decreasing the speed of the reel. and this made i t inconvenient to deal with in the subsequent operations of blending. and.inthelarger sizes. and often does. when. for. afterthe explosion of May.pressed into shape. REELIKG AND CUTTING-UP. in order to keep within the limits of diameter allowed by the Government specification. but it was clear that this operation ought t o be eventually performed by machinery. It was. which supplied the main reason for fitting it with a screw to regulate the pressing. The dimensions of the dies are important. I n the original experiments the cord was wound on reels by hand . any little variation in the system of manufacture or thedryness of the doughmay. and similarly. as i t may now be called.130 On: Tue. as its diameter was increased by winding on the cordite.Proceedings. the size of the die must be carefully watched. and it was intended to fill each to a depth of about 15 inch with cordite. and aIso Delivered by ICEVirtualLibrary. require a variation in the die.] ANDERSON ON THE MANUFACTURE OF CORDITE. 13 Sep 2011 11:38:49 . and especially exactness in diameter.

or he can a t any time stop the reeling for an instant to let a little slack cord accumulate. and the elaborate gear for producing this automatically was dispensed with. as i t made the drying process lengthy and uncertain.consist of two stamped sheetbrass ends with shallow flanges.com to: machine. in slack.intowhich a loose drivingspindlemay be slipped wheneyerthe by ICEVirtualLibrary. however. In the present form of the rifle cordite press. is placed on a separatestand a littlewayin front of the press. as already mentioned i n connection with the original machine. Fig. sticks i t on to the reel. Accordingly theplan of windingonly one cylinder chargeon each reelwas adopted.which nected to the belt striking gear. and is driven by a belt from a cone pulley a frictionon the worm driving shaft.althoughthis involved the use of a greatnumber of reels.was also simplified. reversed like the press-shaft. which was accomplished by means of a pair of cone-pulleys and a belt. 13 Sep 2011 11:38:49 .130 On: Tue. If he should be a little slow i n doing this. and then starts the gear. the reeling gear is not started also. it much facilitated after the operations. All that was found necessary was togivetheattendantthe power of quicklyalteringthe relative speeds of the press andthe reel-spindle to a limited extent. The distribution on the reel is effected by a guide worked by a cam-groove on a rotating cylinder. the variation in speed was found to be practically negligible. Delivered reel isputintothereeling IP: 194. 5 . The pulley is driven by is also conc l ~ t c underthe l~ control of theattendant.Should he find thatthereeling order to gatherupthe is proceeding either too slowly or too fast. The reels as now used. When the press is and is not started again. a handle is provided. but as soon as the cord begins to come out. fixed upon an axle consisting of a piece of brass tube.169. by means of whichthe reel can be quicklyrotated. theattendanttakesthe end.143. 6. to allow for any small variation that might occur. that it was not satisfactory to wind so much on any one reel. [Minutes of a reciprocating guide worked by a cam groove on a rotating cylinder. so that the cord has collected t o any extent. because. thusfilling a depth equivalent to only six or seven layers of the cord.independently of thebeltdriving gear. which distributed the cord evenly over the width of the reel. with such a small depth of winding on each reel.86 ANDERSON ON THE MANUFACTURE OF CORDITE. or less than 3 inch. the reeling gear. I t was soon found. and did not suit the blending system as finally elaborated. Figs. in sucha way that at the end of the pressing stroke the reeling machine is automatically stopped. a turn or two of another handle in either direction adjusts it at once. The reeling mechanism. Plate 2 .

as the speed of emission would be necessarily slower. G. especially as all the small presses were needed to turn out the quantity of rifle cordite for which there was a demand . and could be improved upon. leaving the press-speed the same. it was intended to reel it in a similar manner to the c0rdit. and in the early days of the manufacture. therefore.comwooden stands were Accordingly. Withthis construction the reels can be closely stacked together in the stores or in the dryinghouses. wards cut to method was for some time adopted. SINGLE-STRAND REEL. 87 or when the cord is to be unwound from it. namely. and as an automatic reeling machinefor so many threads would benecessarily complicated. and the Author designed and made a cutting-up machine for the purpose. but latterly thin plain brass sheet has been preferred a t Waltham for the purpose. so thatsix cords could be pressed simultaneously with a 6-inch cylinder. 1 inch = 1 foot. that this plan was not altogether satisfactory. but the flanges are only {-inch deep. was arranged with change-wheels. It was found. The dimensions are the same as those originally used.143.05 inch. and therefore the intermediate presses were called into requisition for the 0.169. The drum of the reel has usually been made of thin perforatedzinc soldered to the ends. authorities the at Waltham decided to adopt a primitive though effective system. As the drying operation lasts for some days. is quired to be cut into lengths of about 11 inches for making into cartridges . This reis F g . the next size in practical use is that having a nominal diameter of 0. in order to allowfor possible variations of speed of emission of the several threads. by which the speed of the reel could be altered to suit this size.About 650 are generally provided under these circumstances. and do the reeling by hand. 9 inches in diameter by 6 inches wide. it is evident that each press requires a considerable number of reels.] ANDERSON OH THE UANUFACTURE OF CORDITE. especially if it is worked continuously day and night as a t Waltham. The reeling gear of the rifle press. to dry it on the reels. a number of to: IP: 194. This Scale. so that in a large factory the reels number many thousands and form a not inconsiderable portion of the plant. and they are both light and strong. however.130 On: Tue.Procccdings. 13 Sep 2011 11:38:49 . Nultiple dieswere used. Leaving now the cordite as made for the rifles. and afterit into lengths.e for the rifles.05-iuch size. which is the most advantageous system to adopt.Delivered by ICEVirtualLibrary.

which can yield if required. and spread themselves outflat. On the side of the roller remote from the press the sides of the groove are cut away. found to be inadmissible . The share of the charge which fell to the portion of each reel was only enough to make a few layers of cord o n i t . For the larger sizes. and having fixed it on the reel. It is shown by Fig. and consists of two narrow flat-faced pulleys. thebelt passes horizontallyalong a groove in the framework. The bottom of the roller isset so as to j u s t touch the tops of the knives as the belt passes under it. and were 14 inches long with no flanges. some G feetapart. on which are riveted transversely a t appropriate distances apart. l1 The lower halves fell when cut on to a sheet of paper under the frame. a p n m e t a l roller is placed over and across it.130 On: Tue. suitable for bundling t. and.15 inch diameter and upwards. CJfinutes of placed round the press. at about the middle of its length. Reeling was. a series of short steel knives projecting about 2 inch from the belt. the requirement was that theyshould be cut into definite lengths. 13 Sep 2011 11:38:49 .com to: IP: 194. Plate 2. leaving the belt exposed. The length of the belt is an exact multiple of the pitch of the knives. and put into a small frame provided with two long worked simultaneously horizontal knives so arranged that they lengthwise of the reel at the ends of a diameter. each having a special reel provided with a handle. The pulley Delivered by ICEVirtualLibrary. wound it up with one handanddistributed it evenly across the width with the other. and at the end of eachpressing.ogether toform the chargefor the cartridges. All were then spread out i n shallow wooden trays with perforated bottoms.88 ANDERSON ON THE MANUFACTURE OF CORDITE.whiletheupperhalves remained on the reel until dexterously removed by the boys. so that the variation which was unavoidable by this process was of no consequence as the limits were not exceeded. the bearings of the rollerare helddown bylight springs. order to prevent any damage to their edges. and ir. carried in a smalliron frame. and consequentlp cut the charge into halves.1 inches. The reels had % diameter of 7. in which they were conveyed to the drying house. 8.169. Between the pulleys on thetop side. theyweretaken from their stands. and has been used ever since for all the large presses. and a boy to turn it. from say 0. and at last the Author contrived a machine which answered the purpose very well. Each boy took the end of a cord as it came out. Considerable latitude wasallowed inthelengths of the pieces. the resulting product being therefore a number of lengths of cordite. averaging when dry about inches. with suitable axles and bearings fixed in a framework of wood or iron.143. and round the pulleys is an endless leather belt 2 inches wide. as already stated.

by which the speed of theknife-belt canbe altered as requiredwithin certain limits. the lengths thus cut up are picked off the belt bya boy. Formerly the nominal standard.] ANDERSON ON TEE MISNUFACTWE O F CORDITE. e corresponding number of belts are required. is limited by the rate which at the boys can remove the pieces from it. with good limits of error.com to: IP: 194. astheleathergave some trouble from stretching on various occasions. length of all the large sizes of cordite was 14 inches. as this is found more practicable than making the knives so that they can be shifted on a single belt. Delivered by ICEVirtualLibrary. as two or three cords can be cut upon it simultaneously as well as one. but latterly several additional lengths have been adopted.and it is also connected to theautomaticbeltstriking gear of the press. Of late the leather beltshave been superseded bythin steel. the knives were placed on the 4 Only one belt was therefore belts a t a pitch of 1 . inches. and the belt moves along towards the roller. A friction clutch is also inserted and placed under the control of the attendant. and 60 feet perminute. i n which is included a pair of cone pulleys and a belt.Proceediugs. and allowing for shrinkage. lying over the edges of the knives. so that the cutting-machine canbestartedor stopped atany time whenthe press is a t work. needed for each press. 89 at this end is driven from the press gear by suitable mechanism. thereby cutting it through theroller to wherethe sides of the groove arecut away. so that the varying lengths of belt necessary can be accommodated.130 On: Tue. so that when the latter reverses for the quick return. carrying the cord on it. and are laid in wooden trayswith perforated bottoms. the cord is forced .143. and an experimental apparatus was constructed for the purpose with moderatesuccess. 13 Sep 2011 11:38:49 . the attendant then starts it. the end is taken and laid on the knife belt . in order to fully equip a machine.The speed of the belt. for drying. it may be operated quite conveniently by hand . For use withdirect-actinghydraulic presses wheretheremay not be any shafting to drive the belt-gear. in whichcase all the speed-regulating mechanism can be dispensedwith. and as itpasses beyond down on to it. As soon as the press is started and the cord begins to come out. and seriously affected thepitch of theknivesinspite of all the precautions taken to stretch it thoroughly before the knives were fixed. The pulley nearest the press is made movable. the cutting machine stops.169. Attempts thishas been found tobeabout were made to make the removal of the pieces automatic. Theapparatus can be used with either single or multiple dies. As each knife comes to the roller. and therefore. but it was found that very little labourwas likely to be saved thereby. and consequently that of the press.

are gathered into an untwisted rope. sixty of the cords. 10 6 9. 74 9. Afterdrying.. have spindles placed i n them. 9. 50 . 93 . 30 .. 9. while the other end terminates witha small weight. it was found that the sticking was experienced. Ten of the reels. i n order to secure the greatest uniformity in the result. 13 Sep 2011 11:38:49 .com to: IP: 194.. but as soon as the final composition and consistency of the material nolonger was established. and no such drying wasnecessary or desirable.. as they come from the drying-house. BLESDISG. and the period of time required for the drying varies according to the size of the cordite.169. G 9.. round which a turn of string is taken. The effect is to producein each reel a slight resistance to turning. Opposite this frame. 5 .143. Each spindle carries a small pulley. 6 v . 33 I.90 ANDERSON ON THE MASUFICTURE OF CORDITE. In theearly experimental stage.130 On: Tue. and in order to overcome this a partial or surface-drying between the press and the reel was suggested . where they are deposited on racks for the necessary space of time.. . G . so that the cords after leaving the press might be dried before reeling. which will be described later. The temperature is maintained by steam-pipes a t about 100" F. in which a systematic mixing of different batches is made.. . Size 15 requires G days.. 0 * 0375 inch diameter.. I . the reels from the rifle presses.. . SO that an average quality may be produced. but. Trouble was at one time found with the strands of cord sticking together on the reels. . [Enutea of DRYING. but this was soon discovered to be impracticable. Thelpresent Waltham practice in this respect is as follows :Size 3 requircs 35 days. in order to keep a slight tensionon the strings as they are wound off'. and are mounted i n a vertical frame with the spindles resting in suitable bearings. thus affording a gentle brake. w 20 >. For filling into cartridges.. 15 ... one end being fixed to the frame. No machinery is involved in this process . continuous drying process was a proposed.the cordite manufacture is completed.. is placed a little Delivered by ICEVirtualLibrary. 5 . and in this condition are fed intothecartridge-filling machine.. 40 . 32 . are transported i n suitable trucks to the drying-houses. and at a few feet from it. andthetrays from the others. the important process of blending has to be carried out.

1 inch = 1 foot. and of the same dimensions. and distributes the threads. againstarted. which contains a reel similar to those already described. with heavybrass endsand deep flanges. the two ends are quickly joined by moistening them with acetone kept handy for the purpose. 7. Should a thread break. 91 machine. care is exercised to draw them from differentbatches i n aproperlyorganised manner. which is known therefore astheten-strand reel. On this. 7. if. and consequently in the shooting. in the selection of the singlestrand reelsto beginwith.130 On: Tue.] ANDERSON ON THE UANUFACTURE OF CORDITE. the ten threads are gradually wound off the ten reels o n to the single one.com to: IP: 194. that. a very perfect average specimen must necessarily result. with a guide fork on the top.169. the ends are passed through the guiding fork and on to the reel of the machine. the drumbeing of tin. The machine is started and stopped easily and quickly by means of a friction driving attachment. the contents of the six ten-strand reels are wound.andthentheten-strand reels from which the completed rope is made. it a reel 8 inches diaSIXTY-STRAND TRANSPORT REEL. It will now be understood. A vertical arm. evenly over its width. ten but larger and having on Scale. Hencethis process of preparing the charge for the rifle cartridges is peculiarly favourable to the production of great uniformity in the ballistics. a statement which has been amply verified by results.143. a& the machine is Figs. A thread from each of the ten reels is now taken. driven by a gut band from the shafting of the factory.Proceedings. Figs. and a very certainand uniform blend be obtained. where they are secured. The sixty-strand reels thus filled with cordite are now ready for storinginthe magazines untilwanted for filling into cartridges. but with a much deeper flange. meterand 22 inches long.Six of the ten-strand reels thus filled are placed in a frame similar to that alreadydescribed. Delivered by ICEVirtualLibrary. is caused to reciprocate in front of the reel by means of a cam of special shape. thus forming the required sixty-strand rope. 13 Sep 2011 11:38:49 . The machine having been started. i n a similar way. which are led in one bunch through the fork. i n front of which is a machine of similar type to that used for the strands. the machine is instantly stopped.

per hour. and then to cut it off. Thefunction of the fillingmachine is to feed a certain length of this rope into an empty cartridge-case. The first proposal was to advance the rope by means of a pair of grooved rollers placed opposite to one another. designed to contain 100 lbs. which is performed by a special machine. 13 Sep 2011 11:38:49 . quick in its action.112 ANDERSON ON THE MANUFACTURE OF CORDITE. by a process of selection from different batches. CARTRIDGE-FILLING. [Minutes of This machinery was designed at Waltham by Mr. i t was to be cut off by a suitable knife outside it. of cordite per hour.130 On: Tue.0375-inch diameter cordite. would move differentamounts for a given degree of rotation of the roller. made u p of sixty strands of 0. Jenkin.theprincipalinterest lies in filling the rifle cartridges. is. so that a very complete mixture is effected. and was made a t Woolwich Arsenal. they are fastened up. but under trial it became a t once evident that differentportions of the groove in the rollers.143.thushaving a circular orifice between them fitting the rope tightly. and simple. an operationwhich is performed twice. They were to be rotated by means of an adjustable feed motion t o the extent required for one charge.com to: IP: 194. I n this section. t o determine the best of method of performing the operation. and as each charge was fed into a cartridge. as already stated. whereas the larger cartridges areall made up by hand. the speed being about 52 feet per minute. As the boxes are filled with completely blended cordite. and consistsmerely in a systematicmaking up of the cut-up pieces in boxes. The charge for the 0*303-inch. Experiments were made by the Author and Sir William Anderson at an early stage the cordite development. so that on turning it up the cordite would fall down inside to its proper position. owingtotheirvaryingradii. The necessary attributes of a machine for this purpose are that it should be accurate in its work and not liable toerrors. The ten-strand re-reeling machines will each deal with 25 lbs. the speed of reeling being about 254 feet per minute. The charge fitted the cartridge case loosely.. wound on a reelin an untwisted rope. asinthelast. A modelwas made. labelled and sent to the magazines until required for making up into cartridges.or magazine rifle. and the effect would beto move the component threads of the rope irregularly.TheAuthorthen devised the followDelivered by ICEVirtualLibrary. and the sixty-strand machine will reel 369 lbs.169. Theblending of thelarger sizes is performed byhandin another house.

are disposed uponabase-plate .Proceedings. slipping over the rope which is held fast by A. then laid in the bottom blocks of the two clamps.whichwasquite successful. therefore.143. Looking atthe machine from the side a t which the operator sits. An empty cartridge-case having been placed in the holder. under the influence of a weighted lever.gain. and communicates with it by a small trap-door so asto minimizechance of accident) is broughtin at the left-hand end through the trumpet-guide. and finally moves back to the left up to the adjustable stop. and the charge is cut off just outside the cartridge-case. so drawing the rope through A. which may be called A and B. The latter is then removed and put into a rack board with its closed end downwards. and the rope of cordite coming from the 60-strand reel (which is fixed in a convenient position outside the house cont a u k g the filling machine. Two clamps. again both clamps are gripping the rope. after which comes a cutting-knife operated by a hand-lever. To the left of A there is a fixed trumpet-shaped guide. theragged ends of the rope aretrimmed ~ f by a cut of theknife. and the machine is again in the normal position ready for another operation. A is fixed. The machine is worked by a pedal. and the first action of the mechanism is to close the clamp A .com to: IP: 194.he pedal next opens the clamp A. As soon as the correct stroke has been made. while B is capable of making a reciprocating movement by sliding in a groove formed inthe base-plate. and to the right of B there is also a fixed guide. whichisnormally held open. in the sense thatitmay be made to advance towards or recede from A. at this moment. Delivered by ICEVirtualLibrary. B is placed to the right of A.andthe f machine isreadyto work. and then moves B (gripping the rope all the time) to the right. the operator presses down the pedal . the knife-lever is pulled over. A further movement of t. which is determined by adjustable stops. and finally a cartridge-case holder.169. which.fits i t loosely.] ANDERSON ON TEE MANUFACTURE OF CORDITE. and its operation is as follows :-The top halves of both clamps are turned back. The clamp B is now opened. A t this moment bcth clamps are closed because A is always normally closed. upon which the charge of cordite. 93 ingplan. The halves top of the clamps are then turned down a. andisthearrangement used up to the present time. and finally passed through the right-hand guide. By this time the pedal has returned to the top of its stroke. asalready mentioned. the movement of the pedal ceases . the first result being thatclamp B. and pushing it through the righthand guide the into cartridge-case beyond. The pedal is now allowed to rise. drops tothe bottom. 13 Sep 2011 11:38:49 .130 On: Tue. closes andgripsthe rope.

so that when the clamp is closed. [Minutes of The clamps act in the opposite sense to one another. is perforqed bypressing the pedaldown and allowing it to return. and so frees the rope . guided to move in an approximately verticalline. allowing theclamp B first to close. while that of the movable clamp B. when one is open the other isclosed . and can thereforeneverslip or getout of place.130 On: Tue. rises throughanopeningintheplateunderthe lever of the clamp A . thatis. The whole operation. assisted by a small lead weight at theoverhangingend remote from the hinge. so that in this position both clamps are free to grip the rope. The frame is connected with a vertical rod coupled to the pedal lever. so that as i t is depressed the strip lowers and the peg rises. two such pegs rise. As the top lever is raised this hole is enlarged. except the cuttingoff. one from each end of theframe. The pressure for givingthegripisderived solely from the weight of the top lever.but there is always a brief space of time during the process of reversal when both are closed. while. opening i t sufficiently to allow the rope toslipthrough.com to: IP: 194. and it is carried by bearings attached to the base-plate. the action issimple. then maintaining both closed for an instant. Underneath the base-plate is a rocking-frame. SO that the rope is always held by one or the other or by both. transverselyo the laminat tions. the leaves i n one corresponding with and entering into the spaces in the other. unless the lever is turned well back. overhangs on the side next him. afterwards raising the lever of the clamp A. 13 Sep 2011 11:38:49 . Thelever of the fixed clamp A overhangs the base-plate on the side remote from the operator. and exactly follows the pedal movement . but the laminations are sufficiently deep that in working they always interlock with one another.91 ANDERSON ON THE XANUFACTURE OF CORDITE. hinged one side at of a laminated bottom block . Plate 2. and thus the strands of the rope can never get out of place. Fromthisframe a peg.neitherthesingle peg nor the strip are incontact with the levers. When this frame is horizontal. and therefore always form an enclosed opening. from the opposite side.143. SO that as far as the operator is concerned. there is in effect a round hole formed right through it. which the centre fulcrum spindle of lies directly under and parallel to the axial line through clamp the openings.169. which lies under the lever of the clamp B. of a size to tightly grip the rope of cordite. but are a little clear of them .The object of thelongstrip Delivered by ICEVirtualLibrary. The clamps consist each of a laminated top lever. The clamp B is attached to a plate which can slide about i n a longitudinal guide-groove in thebase-plate. The machine is illustrated by Figs. V-shaped notches with rounded bottoms are cut through both. 9.joinedatthetopby a smooth flat strip.

and is capable of a definite amount of movement in it .Proceedings. the top of the slot comes into contact with the block. I t remains to describe how the reciprocating movement of B is effected. and now thebell-cranklever beginsto move. were apt to buckleupatthe conlDelivered by ICEVirtualLibrary. and therefore no motion of the clamp-slide at first takes place.143.130 On: Tue. i n order to insure that none of the strands catch the edge of the latter while being in and on fed so perhaps damage the charge. or other cause. 13 Sep 2011 11:38:49 . the backward movement commences.169. but as soon as the bottom of the slot reaches to the block. the block is at the bottom of the slot in the rod. I n orcier to prevent the closed clamp from being accidentally pressed down too far.com to: IP: 194. by a hand being laid upon it. While is being moved forwards and backwards. the longer vertical arm of which is connected by a link with the sliding plateof the clamp B. the lever the slide frame controlling theclamps is still moving also. carrying an oblong brass slide-block. it offers a slight resistance to the passage of the rope. Under the base-plate bracket a projects downwards. andasthisguide has to smaller be than the opening of the cartridge-case. In thefirst machine a difficulty arose whichisworthy of mention. which allow them to be closed only sufficiently tightly to grip and no more. andthe slide is drawn forward till the right-hand stop is reached. that is tosay. owing to the clearance between the block andthe slot. which are of small diameter and therefore incapable of offering much resistance tothrust. to the clamps are again reversed before any motion of the slide takes place. when the pedal is first pressed down. thattheindividual cords. The shorter horizontal arm has a pin in it. On allowing the pedal rise. but provision is made so that the unnecessary extra opening of the clamp which thus occurs is of no consequence. The rope is pushed by the clamp B through the guide totheright of it. The action may now be clearly seen . 95 under the lever of the clamp B is that it can act upon it in any position in which it maybe situated.] ANDERSON ON THE ILIANUFACTUREOF CORDITE. by which means a small amount of artificial resistis given it. adjustable stops are provided on the levers. The sliding-plateof the clamp B has a spring-brake attachmenton the under side. having a fulcrumpin atthe end carrying a bell-cranklever. which works in a slot formed in the vertical rod coming up from the pedal to the rocking frame. and so damaging or perhaps even shearingthe rope. Theresult is. but after the clamps have been reversed.the slot isabout 1 inch longer than the block. which to and prevents any unance motion to necessary movement from taking place.

[Minutes Of mencement of the stroke and get caught between the clamp and the guide.130 On: Tue. namely. it was desirable to find another solution of the difficulty in subsequent machines : first. was a small stop projecting up into the groove of the slidingplatealready mentioned. for twoimportant reasons.wide from ever exceedingamaximum length. because making two strokes to each charge increased the liability to errors. and guide mould all be closed together. underneath theclampslide.equivalent to a little Delivered by ICEVirtualLibrary. Whenthe clamp wasinitsextreme position at the left-hand side. it might either stay where it was. that is. the distance between the fixed stop and the end of the groove was increased. according to how far it was held or not by itsown friction. and secondly. which corresponded with the ring taking up a mid-positionbetween theclampandthe guide. This.96 ANDERSON ON THE MANUFACTURE OF CORDITE. and this wasfound toanswerwell and to completely preventany buckling of the cords. and was connected with a flat foot lying in an oblong groove on the under side of the plate. It consisted of a brass ring loosely encircling the rope carried by a small square pillar. because it occupied more time. so arranged that it floated between the two. but no satisfactory solution was arrived at until the Author proposed a very simple device which answered perfectly . I f it remained where i t was. The only remedy that could be adopted inthe first machine was to make the charge in two strokes. As the clamp moved to the right. If it moved it would go up as far as the guide. and the ring foot could then take up any position between the two. Many suggestions and some experiments were made. and therefore the foot was constrained to occupy this position at thistime. or move forward withthe slide altogether or in part. and there it would necessarily have to stop. a movable intermediate support for the rope. a consideration which would tell seriously whenthe process came to worked be on a large scale. but never left on either side of it a greaterunsupportedspan of rope thah half length the of the charge.143.com to: IP: 194.then. so completely spoiling the charge. the distance between the stop and the termination of the groove was a little more than the length of the ring foot. Attached to the base-plate. clamp. placed between the clamp B and the guide.alwayspreventedthe unsupported parts of the rope between the clamp B and the right-hand . At the end of the stroke the ring. the clamp would soon come up to the ring and carry it o n with it. 13 Sep 2011 11:38:49 . But.169. which passed through a slot in the sliding-plate of the clamp B. and would not be unlikely to lead to half charges being occasionally put in. thus leaving only half the former length of rope unsupported at beginning the of each stroke .

which have been already capped. and thus obviated all trouble from buckling. by a carefully organizedsystem of inspection. and as at Woolwich. an attachment was provided which automatically locked the knife-lever until the full stroke had beeu made. thisattachment not inthelater machines.P~oceedings. C. made like a chimney and passing up through the roof. A small trapdoor in the side communicates [THE INST. 97 over half of the full stroke. and the length about I&inch. and so no further steps were taken in thisdirection. however.herefore. as it is possible to make a short stroke by not B bringing the clamp quite upto the guide. covered with a light flap to exclude rain but capable of permitting the escape of the gases should the reel by any chance catch fire. The weight of cordite in each charge is about 30. are placed bottom upwards in the trays.com to: VOL. which severs the cords more cleanlythan a direct cut. 13 Sep 2011 11:38:49 .130 On: Tue. and so arranged that a drawing cut is given when the hand-lever is pulled over. As. The empty cases. and communicate with it by means of small trap-doors through which therope of cordite passes.169. this was foundtointerfere somewhat withthe speed a t which the machine could be worked. but the Woolwich authorities seemed to think that this would not be necessary or perhaps desirable. theremainder of the space being utilized for the accommodation of the trays for empty and loaded cases. The was removed. CXXXII. and t. In some instancesthe reel is enclosed i n a sheet-iron casing in the house itself. After a little practice a boy can fill 6. It will be observed that there is achance of error i n filling with this machine. The speed a t which the machine can be worked is considerable. Thecircularknifeadmits of being turned round. the reels containing thecordite are placed in compartments outside the filling-house. The apparatus as described is fitted on a wooden table 2 feet 6 inches square.1 ANDERSON OH THE NANUFACTURE OF CORDITE.143. Delivered by ICEVirtualLibrary. in some of the earlier machines.a false charge would be almost certain of discovery. and occupies the left-hand corner next the operator. As already mentioned. Various ideas were entertained of making the machine more automatic and evea of working it by power.600 cartridgesperday of 8 hours. and as they are filled they are reversed.2 grains.] R IP: 194.E. and was fitted knife consists of a circular piece of sheet-steel sharpenedall round the circumference. and so prevented the charge from being cut off. and consequently the boys would beobliged to be verycarefulintheir work.Allsubsequent machines have been fittedwiththis arrangement.so that asone part becomes dull a fresh piece of the circumference can be used.

and it is surprising how accurate the weight of the charge usuallyis. OF Cnfinutes of with the filling-machine. on reaching the bridge the gun-cotton instantly burns through. the recessed plate :-A is drawn out and a handful of wads is thrown on to it. which canbe drawn horizontally in and out of the frame. i n order t o prevent it from extending through the trap-door to the reel.com to: IP: 194. The tubes communicate with the hollow interior of the plate. corresponding i n pitch with the cartridge cases as they are placed in theholes of the trays used with the fillingmachines. a safety device is provided consisting of a heavy brass guillotine door capable of closing the small trapdoor. The first arrangement is. 13 Sep 2011 11:38:49 . however. which thoughperby haps hardly forming part of the present subject. which a t once falls. and a large door is also provided giving access to the reel. an inspector looks over the trays of cartridges and selects some a t haphazard from eachto test. The errors thus discovered are few and far between. cutting through the cordite rope. opening.98 ANDERSON ON THE DIARUFACTURE CORDITE. and having on its upper side a number of shallow circular recesses of suitable dimensions to contain each one wad. the better of the two. releases the door. and so preventsthe fire from spreadingtothe Although these precautions are rightly taken. under which the cordite rope is made to pass. Below there is a space into which the trayfull of filled cartridges can be slipped. Should the cordite fire.143. Cardboard wadsare next inserted machine. which he does by removing the charge and weighing it in a delicate balance.which by Delivered by ICEVirtualLibrary.130 On: Tue. their positions when the plate ispushed in corresponding with those of the tdms in the upper plate. After filling. and closing up the reel. it is but fair to say that they seem almost superfluous. but which is usually heldopen by a cord passing over pulleys and attached by means of a piece of gun-cotton to a small wooden bridge near the filling machine. which i n its turn is connected by a flexible pipe and cock with a small vacuumpump. and fitted on its under side with a number of short projecting pieces of brass tube. afterwards replacing in the case. as the reels need never be brought inside the fillinghouse.169. It consists of a frame having at the top a horizontal plate capable of being moved vertically through a certain range by appropriatemechanism. Theactionisas f o ~ ~ o w s tray of cases having been inserted. . is well worth a brief description. as the Author believes that no case of such an accident as they are designed to cope with has ever yet arisen. I n case thecorditeshould from any cause catch fire at the filling machine. Under this plate is another.

I n front of this frame a disk is rotated at a speed of about 800 revolutions perminute. such as those known as '' quick-firing.Proceedings. which thus slice up the cords as they come through thin into transverse wafers. tied with string and fitted with the appropriate primer. whereupon the cock of the vacuum-pump is opened. each being nominally 0. and t h e cords are then bundled together into the form of a cylinder.130 On: Tue. t h e cases. after which the first one is removed and emptied. The plate being raised again a little the sliding plate is withdrawn. leaving the wads attached to the ends of the tubes. The charge for the large guns is made u p i nseveral lengths.] ANDERSON ON THE MANUFACTURE OF CORDITE. namely thin wafers.2 inch diameter. The charge of cordite for all the largersizes of guns is made up by hand entirely. Preparation of the Charge for Blank Cartridges. Finally the tube-plate is lowered. whereby the wads are pushed into the cases waiting below to receive them . Thecuttingthendrop throughthe base-plate on which the machine is fixed into a receptacle below.000 cartridges in S hours. whilethetube-plate This machine will wad 75. The quantities required are weighed out. and upon breaking the vacuum. A measured quantity of the cuttings is filled into the Delivered by ICEVirtualLibrary. The cartridges are thennecked and bulleted. and immediately the wads stick themselves on to the ends of the tubes.0055 inch and the quantity turned out per hour is 12 lbs.169. sometimes supplemented with a short tube of cordite. When a receptacle under one shoot is full. 99 rubbing them about with the hand soon fill all the recesses. a slide is moved.143. for which purpose two shoots are arranged. Those for the blank cartridges are made i n a machine designed by the Government engineers. they detach themselves from the tubes andremain in is raised againtoits position. I f for guns having brass cartridge-cases. consistinggenerally of a small bag of gunpowder.plate is pushed into the frame. consisting of a vertical frame perforated with a number of holes. The upper plate is now brought down until the tubes touch the wads. but if for others i t is enclosed i n a suitable bag. er i n some cases a truncated cone. one on each side of a vertical centre diaphragm plate. and into either of which the stream of wafers can be directed.carryingfour cutting knives. and the othershoot comes into operation and diverts the cuttings into a second receptacle. through which are fed by a system of grooved rollers a number of strings of cordite. and the. 'The superfluous wads are then swept off. 13 Sep 2011 11:38:49 R 2 . a different form of the finished product is required for this purpose." the bundle enters the case as it is .-As mentioned at the commencement of the Paper. The thickness of the wafers is 0.com to: IP: 194.

the walls. A place (not shown on the drawing) must thereforebeprovided where the shoes canbe put on. 10. in order to avoid bringing any grit into the place. it is possible to explode the material by impact . but also to prevent any chance of such pieces being trodden on by boots with nails inthem. few requiring any regular foundations. is the engine-room. and it is therefore important. It also drives a small set of three-throw pumps supplying an accumulator. but not built into. that there is reason why the cordite manufacture no dust. therefore. It is true experiments have shown thatwhensuchan explosion does take place. The floor may be made of wood covered with linoleum. which might possibIy cause an explosion. and most of the machines may be simply bolted down to it.100 ANDERSON ON THE MANUFACTURE OF CORDITE. attached to. It consists simply of a long brick wall not less than 14 inches thick. A convenient arrangement of the machinery i n a nzoderate-sized factory is given in Fig. or walking about there in nailed boots. containing a suitable engine for driving a line of shafting running overhead through all the cells. for although the manufacture is very safe. it might besufficientlyunpleasant. I n the event. not only to keepthe whole place scrupulously clean and free from pieces of explosive lying about. Plate 2. in such a way that the brickwork projects well above and beyond the woodwork. 13 Sep 2011 11:38:49 . Outside the cell doors there is a clean wooden platform joining them. Eachcellhas a wooden lean-to roof and frontwith a door and windows. The end compartment on the right. as lightas possible. which provides the necessary high-pressure water for to: larger press. The reason for this precaution is obvious. [Afindes Of cartridge-cases by a simple apparatus calling no special descripfor tion.One is so safe is. Accuracy is of course of no great importance for blank charges.130 On: Tue.169. as Delivered by ICEVirtualLibrary. No person is allowed on this unless provided with suitable clean leather shoes. which forms such a terrible source of danger in black powdermaking. with a series of partition walls of %inch brickwork running out from it. of an explosion occurring in any cell.com the IP: 194. 10. and the walls would most likely remain sound and prevent any spread of the explosion. and raised above the ground. the front and roof could be blown out without doingmuch damage. dividing it into a number of cells. andwherethe workmencanchange their clothes. but even so. and does not spreadbeyond the point immediately affected.143. GENERAL ARRANGEMENT OF TEE MACHINERY IN A CORDITE FACTORY. Fig. it is invariably a local one.

if charging rammers are used.” that is. and must in England be installed in accordance with the Home Office regulations. Besides the machinery-house. blending. After this two compartments. so as not to leave open communication between the cells which might assist the spread of fire.143. as will beseen by figures already given. there is a cell containing either a direct-acting hydraulic press.] ANDERSON ON THE MANUFACTURE OF CORDITE.130 On: Tue. The compartment next the engine-room acts as a service store for paste. C. 101 well as the charging rammers.E. If an extension of the factory is required. &C. which will probably prove sufficient in practice to deal with all the rifle cordite produced. its output being nearly 50 per cent. 13 Sep 2011 11:38:49 . if all four presses are continuously worked on rifle cordite to their full capacity. Delivered by ICEVirtualLibrary. Minutcs of Proceedings Inst. or else an intermediate screw-press. in either case provided a set of apparatus for hand with a beltcutting-machineand reeling and cutting such sizes as 0. mi~eral jelly and acetone. the dough that remains in the bottom of thecylindersafter each charge. a similar set of cells may be built on the other side of the main wall. Jenkin . cx.169. The machinery-house may be warmed by the cxhaust from the engine. several otherbuildingsare required for drying. greater than that of ten-strand. each with an incorporating machine. F. The wall-boxes in the partition walls through which the shafting passes are provided with diaphragm plates of‘ iron closeiy fitting the shafting. though. The boilers are placed separately a t a proper distance from the danger buildings. Finally there is a cell containing one ten-strand and one sixtystrand re-reeling machine with their stands. Next.Proceedings.05 inch. and a second one would be required.. but itshould be remembered that they do not necessarily comply with the Home Office rules.theten-strand machine willhardly be able to keep pace with them. vol. size. There is plenty of margin in’the sixty-strand machine. p. Then come three cells. C.com to: IP: 194. two being of the 150-lb. 367. but these do not properly enter into the scope of the Paper. and steam for driving the engine and for warming is conveyed to the several houses by long pipes. with a second line of shafting through them. one is placed in each compartment to serve the two presses contained therein. Electric lighting is the most convenient to adopt.. and. each large enough to accommodate two rifle-presses and their accessories . The electric lighting arrangements of the Waltham factory have been described by Nr. and the third a small one capable of being used for re-incorporating the waste material or “heels.

ThePaperis accompanied by eleven drawings. . [IIinutes of from the operation of which theGovernment factoriesare excluded. The Author is indebted to Colonel Ormsby. the Superintendent of the Waltham Abbey Factory.and other buildings a t Waltham are not necessarily such as private manufacturers would be at liberty to adopt. from which Plate 2 and the Figs. it is believed. For the same reason the arrangements of the machinery.169. in the text havebeen prepared.143. The building which has been described would.com to: IP: 194. though in some cases local circumstances may necessitate modifications. generally receive the approval of the Home Office.130 On: Tue. Delivered by ICEVirtualLibrary.102 ANDERSON ON THE MANUFACTURE OF CORDITE. 13 Sep 2011 11:38:49 [ l)rscussloN. for much of the information contained in thePaper.

143.130 On: Tue. 13 Sep 2011 11:38:49 .com to: IP: 194. W ANDERSON Delivered by ICEVirtualLibrary.E .169.