Part 21

Fig. 299 is a section of the chuck which carries the rosette, the latter being shown in position. A, is the body of the chuck; B, the inside screw to fit the mandrel. On the outside of this part are a few turns of a somewhat finer thread, beyond which a plain part is left next to the flange, C, and on this the rosette is placed, and is clamped by the nut, cut outside into beads, or milled, or left plain and drilled for the insertion of a lever pin; this nut is marked D in the figure. It will be seen to lie within a recess formed in the face of the rosette, E. The latter is shown in Fig. 300. A, is the recess just alluded to; B, C, the pattern on the face and edge; D, the large hole in its centre, allowing it to be slipped on the back of the chuck. It is prevented from turning on the latter by a pin, which fits into the small hole, F. From the front face of the chuck rises a conical pin, G, similar to that on the eccentric chuck, over which fits the circular division plate, I, with its projecting screw, H, to hold the ordinary chucks. This plate is recessed at the back, leaving a mere ring of metal, _e, e_, which fits a corresponding circular groove, and secures the steady movement of the plate, which is fixed by a conical washer and screw, as seen in the plate. This division plate is formed with cogs, into which a stop falls, as described, when treating of the eccentric, but the divisions are differently arranged, the cogs being divided into sets of eight or ten teeth. Let the whole circumference be first divided into six equal parts, and, beginning at the first division, cut six or eight teeth, as if the whole circle were to be divided into 60. Pass to the second portion, and cut eight teeth, as if the circle were to be divided into 72. Let the third carry eight teeth, with a pitch of 80 to the circle; the fourth a similar number, with a pitch of 84; the fifth 96; the sixth 100 to the circle. These must be cut with the same cutter, so that the spring click, or stop teeth, may fit any of the six sets. There will be several undivided spaces, which are to be left plain. The several sets are to be marked in numbers, so that the pitch may be discovered at a glance, and the teeth in each being so few will hardly require separate numbering. The part of the apparatus which carries the rubber--that is to say, the heads of the branched part of the swinging frame, are made flat on the top, which projects on three sides, forming small tables on which the actual holder or clamp can be fixed by a turn of the screw, X, in the first figure. The rubbers are merely flat pieces of steel, with edges sufficiently sharp to penetrate to the lowest depths of the undulations and recesses on the rosettes, but rounded off and polished, so as not to cut and damage the softer metal against which they act; other materials have been tried, as ivory, and the harder tusk of the hippopotamus, but hard steel is most generally preferred. It is, of course, necessary that the rubber press with some force against the rosette, which force should, moreover, admit of being regulated at pleasure. This is effected by a spring of steel under the lathe bed, to the end of which is affixed the arm, W, which has a ring handle nearest to the operator, and is perforated with a row of holes from end to end. This arm is flat, and falls into a fork at the end of the tail piece, Y, which is seen in the second figure attached to the centre of the lower bar of the frame. A pin passes through holes in this fork, and through one of those in the arm. By this arrangement it is easy to regulate the force which the spring shall exercise, as this will be increased by moving the pin nearer to the spring, and diminished by placing it in a hole nearer to the ring handle. In forming the rosettes great care should be exercised to make the corresponding parts agree. The depth of similar hollows must be precisely equal, and elevated portions intended to match, must do so with great accuracy. Supposing, for instance, a rosette to be made with ten elevations and ten recesses, all of equal curves. If these are accidentally unequal, and it is desired to arrange a set of these waved rings one within the other, so that the depressions of the one shall be opposite to the elevations of the other, or so that this effect shall take place gradually--if the curves of the pattern are unequally cut the several portions of the device will not tally, and an irregularity will be produced of disagreeable appearance; an inspection showing at once that such irregularity is not part of the device, but unintentional and erroneous. This leads to a consideration of the division plate of this rose engine, and an explanation of the object of its peculiar construction. The pattern of the rosette is, as it were, in sections; either similar elevations and recesses are alternately repeated, or there may be a variety of such, extending over a part of the circumference, and forming a certain complete device, which may then give place to a second pattern, extending a similar distance; and these two may alternate regularly round the circumference. Each of these sections must be of precisely similar length, and the repetition of the pattern must also be precisely similar, for the reason stated above.

It is evident that all possible alternations of the device or set of devices may be obtained by means of six or eight notches of the division plate; for, by moving it forward to that extent, the whole pattern may be compassed. In treating of the second form of rose engine, and of the method of using it, this will again be adverted to, and illustrated by an example.

Hitherto we have spoken only of the use of the pattern on the edge of the rosette, but, as already stated, it is frequently repeated on the face, and by this means it becomes easier to work upon the outside of a cylindrical piece, as well as on the two ends. Patterns thus cut on the face of a rosette tend, of course, to move the mandrel to and fro in its collars, which is only possible when the lathe is made on the plan of a screw-cutting or traversing mandrel lathe. The spring which keeps the rubber in contact with the face of the rosette is shown at P.

The head on the top of the arm, which forms the clamp or holder of the rubber, must be turned round to face the rosette, or a separate rubber must be used, which passes through the clamp at right angles to that used in surface work, and the frame must be prevented from oscillating by a stop which fits between the bearers of the lathe, and embraces the upright side of the frame. The top of the slide rest must be turned round, or a side tool used. In this case, the rest and frame with its rubber become fixtures, as in ordinary turning, the mandrel and work being alone moved in correspondence with the pattern on the face of the rosette. It would indeed be much better to do away with the frame altogether, fixing the rubber to an upright pedestal mounted on the lathe bed, and using the slide rest in the usual way, were it not that in general the work is constantly being varied, the side and face being worked in turn--and the apparatus is rather cumbrous to remove and re-mount often. It is, nevertheless, easy to finish the ornamentation of plane surfaces first, and then to remove the frame altogether, and substitute a fixed rubber, as stated.

The rose engine now described has certain evident advantages over the rose cutter frame, and is capable of the most exquisite devices. It may be, perhaps, a mere question of taste, whether rose engine work finely executed is not in point of beauty, superior to any that can be done by geometric chucks, however elaborate. That it is so, is decidedly the writer's opinion, more especially when this apparatus is used in combination with the oval chuck. Moreover, there is nothing in the form of rose engine described to make it a very expensive article, or beyond the skill of an amateur; and with a set of only three rosettes, the patterns may be varied continually, and multiplied--if not _ad infinitum_--yet quite sufficiently for the display of skill and taste of the operator.

In using the rose engine, it is necessary to carry the cord to the pulley from a very small wheel on the axle. Sometimes the lowest speed pulley of the flywheel may answer, but if the recesses of the rosette are deep and sharp, or only slightly rounded, it may become necessary to mount a still smaller wheel on purpose; else the rubber will jump over and miss parts of the design, thereby spoiling the work. The watch-case turners, indeed, altogether dispense with the flywheel, and use instead a small pulley, fixed to the side of the lathe bed, and turned by hand.

These artificers always use a more elaborate form of rose engine, which will be presently described, and which is the most perfect in detail of all similar contrivances, but it is necessarily costly, and cannot be said to be well adapted for plain turning also, except in a limited degree. The work, if not of soft material, like a watch-case, should be turned first of all upon an ordinary lathe, the mandrel screw of which is a counterpart to that of the rose engine, and the latter should merely be used for the final cut, to perfect the form of the material previous to its ornamentation.

In the Appendix will be found a new method of obtaining the required oscillatory motion of the rose engine, which might apparently be applied to tool holder[22] frame here described, or to the poppet head.

[22] _I.e._, the main frame carrying the slide rest.

The rose engine proper is arranged with an oscillating poppet head carrying the mandrel and its rosettes, the tool being stationary. The following account of this machine, and the drawings, are copied almost exactly from Bergeron. The modern rose engine is not indeed made with the projecting lugs referred to as intended for the application of the guide ring in oval turning, as this guide is now altered to fit a poppet head of ordinary form, as already detailed. The pulley and division plate are also of obsolete form, but as the main arrangement of parts described are sufficiently similar to that now followed, Bergeron's drawing and description have been retained. Fig. 302 is a longitudinal view, and Fig. 303 a transverse view of the working parts of this lathe. A, A are the poppets, which are in one casting, with the connecting piece shown by the dotted lines, which latter has a tail piece firmly attached to its centre, to which a spring is affixed as in the lathe previously described. In the drawing the cylindrical collars carrying the mandrel are split, so that in case of wear they can be tightened in the usual manner by a screw at the top of the poppet marked B, 303. The lugs, _f, f_, with square holes _e, e_, are for the application of the guide for oval turning, the latter being originally a ring with slotted arms on either side. The points of oscillation are precisely similar to those of the rose engine first described, two short poppets C, Figs. 302 and 303, having centre screws, whose points fall into conical holes made in opposite faces of the poppet, a little below the level of the lathe bed. These are formed with a slit to receive the stop _h_, which is hinged at the point _o_, and which, when raised by a wedge, catches into a small projection _p_, thereby fixing the poppet in a perpendicular position and preventing its oscillation. The rose engine can then be used as an ordinary lathe, to finish the preparation of the work to be operated on, which should, if possible, be commenced and mainly formed on an ordinary lathe, the mandrel of which is a counterpart of that of the rose engine. The tail piece, E, does not require a separate description, being precisely similar to that already described. The to-and-fro movement of the mandrel caused by the action of the rubbers on the face of rosette, is also arranged in a manner similar to the last. F is the spring, turning in the middle of its length on a pin in a piece of iron fixed on the bed, so that if both ends wore free it could swing backwards and forwards between the cheeks of the lathe on this pin as a centre. The upper end of this spring is branched in a semicircular form to embrace the mandrel, this fork falling into a groove formed to receive it. It can thus be brought to bear against either of the shoulders visible at this part. The lower end of the spring fits into a notch, or rather a slot in the arm H of the second figure; the handle of this arm being L in both figs. This piece is pivotted at K, and at its other end falls into one of the notches in the retaining plate, G, of the first figure. By this plan the tension of the spring, can be brought against the mandrel in either direction at pleasure, for if the lever is placed in one of the left hand notches, the tendency of the spring will be to move the mandrel towards the right, and _vice versa_. The tension of the spring can also be regulated by the use of the groove B or X at pleasure. In the second figure of Bergeron's the piece which at first sight appears to be a continuation of the holding down bolt of the short poppet carrying the centre screw, is the tail piece or lower end of the long spring just described, and its reduced extremity is visible passing through a short slot in the lever H, near the handle of which appears the _edge_ of the notched piece G of the other figure. All the above parts are commonly of iron, the following are in brass or gun metal.

On the bed and parallel to it, two pieces of brass, or standards, rise, similar to H in the first figure, the two being opposite to each other, one on each side of the mandrel, as shown in the second figure. Both these are firmly secured to the bed by long bolts and nuts, it being of the utmost importance that they should not move or vibrate in the least. They are in addition united to each other by two horizontal braces, one of which is seen at N in the second figure. At _l, l_ are seen two rectangular notches, which are the ends of grooves made in the upper part of the head piece, H, and which traverse its whole length. They receive the crooked part, _a_, of the rubber holder, Fig. 304, so that the latter can be slid along this bar, and brought opposite to any one of the rosettes, after which it can be secured in position by the screw, _b_, Fig. 304. The mandrel is thus arranged. It is cylindrical, with a shoulder against which the chucks can rest, as in an ordinary traversing mandrel, and a similar but reversed shoulder at _g_, Fig. 302. Against the latter abuts the end of an accurately turned sleeve of brass, which fits over the mandrel with slight friction, so as to have no shake or play upon it. Upon this sleeve the rosettes are placed. They fit accurately over it, and are prevented from turning round upon it by a feather extending the length of the sleeve, which fits into a corresponding notch cut on the inside of the rosettes. These are arranged in pairs back to back, and each couple is separated from the next by a short sleeve or ferrule, which Bergeron recommends to be of wood, as tending to hold the rosettes more securely than metal when pressed together by the nut at the end of the set. The fibres of the wood are to be placed parallel with the mandrel, because there is no shrinkage of this substance as regards its length. The pulley is fixed beyond the rosettes on a part of the mandrel filed into six faces for that purpose, and lastly comes the nut, which secures all the parts to their several positions, but which nevertheless does not so jam them together but that the mandrel can be turned within the sleeve when the positions of the rosettes are to be changed in the course of working a pattern. The division plate is not attached to the pulley, though lying close upon its surface. It slips on to the sleeve on which the rosettes fit, and its spring-catch only on the face of the pulley. Thus the latter is held to the sleeve and its fittings when the catch is down, so that all turn together, but, when the catch is raised, the division plate, carrying with it the rosettes, can be turned round upon the mandrel as may be required. It is not necessary to repeat what has been said respecting the manner of graduating the division plate, as that used with the lathe already described is in that respect a counterpart of what is used with the rose engine now treated of.

The following description of the method pursued in turning a pattern shown in Fig. 307 will suffice to show the working of the rose engine:--First, says Bergeron:--It is not enough to know the general construction of the rose engine, it is necessary to know thoroughly the particular one in use, _i. e._, as regards the details of its construction, the slight defects or imperfections it may chance to have, and the means whereby they may be lessened or corrected. It is necessary, in addition, to know well, and to have always at hand, the numbers of each rosette, or any rate to have a table of them which can be readily referred to. It is equally necessary to recognise at a glance the various sets of divisions on the division plate, for which purpose, and that no mistake may be made, such numbers ought to be engraved upon each. The same holds good with regard to the slide rest, and, in addition, practice should be frequent upon box or other inexpensive material by which the turner may have made himself perfect in the several combinations possible, and the various effects producible by the rosettes and different shaped tools over which he has control. It is thus, by actual experiment only, that the turner can become acquainted with the powers of his own lathe and apparatus, and thus only, after working out the patterns already executed, will he be in a position to design new ones, and to work with ease and certainty. The rose engines are usually fitted with tools of variously shaped edges, as shown in Figs. 305 and 306, by this means a pattern of some width and great variety is of course produced at once, and by one rosette. In the following, however, a tool with single point, Fig. 308, is to be used. This simplest design is supposed to be on the cover of a box or other plane surface, and it is evident that the movement or oscillation required of the mandrel is that at right angles to the bed of the lathe. To obtain this movement, when the rubber is fixed in its clamp, on the side of the workman, as it is necessary that the rubber should press against the rosette through the medium of the spring, the handle of the lever, Fig. 302, must be drawn forward towards the operator, and kept by a pin, as described, passing through it and the tail piece of the mandrel frame. The tension must not be too great, especially if the rosette to be used is deeply indented, and care must be taken to free the frame from the action of the stop, _p_, by removing its wedge before making any attempt to try the pressure by moving the mandrel. The design under consideration is produced from rosette numbered 2 in the drawing, and in fixing the rubber care must be taken that it does not bear against the adjacent rosette. Choose a rosette of forty-eight teeth or undulations, and as the second circle of ornamentation exactly intersects the first, the raised part of the one falling under the depression of the other, and as it were _halving_ it, the set of divisions on the click plate to be used will be twice 48, or 96. Place the rest parallel to the face of the work and so that the forward motion of the tool shall be perpendicular to it. By means of the leading screw of the rest, place the tool near the edge of the work and level with the centre, and gently moving it forward and putting the lathe in motion, commence the cut. After having made a light cut, without moving the tool, stop the lathe and judge of the depth of cut, and if sufficient, screw up the stop screw of the slide rest, to insure all the following cuts penetrating to the same depth. Observe the position of the tool as marked by the graduations of the slide rest, and then withdrawing it from the cut, move the click plate one notch, which will divide exactly in half the several undulations of the rosette. By the rest screw move the tool towards the centre of the work and mark the number of divisions passed over, so that the circles of undulations may be equidistant, and cut a second. Now for the third, _go back or advance_ on the click plate one division, for the position of the undulations in the third is precisely that of the first circle. It is indeed immaterial whether an advance or retreat of one notch is made in this case, but now is evident the reason for not dividing the plate equally all round, five or six teeth being ample for each division. If there are eight rosettes the plate should be first divided into eight parts, and each rosette having a different number of undulations, these eight parts should be divided into degrees proportionate to the numbers on the rosette, the one being a multiple of the other. In working the _side_ of a cylinder, that of a box, for example, the longitudinal movement of the mandrel is required, the poppet being retained immovable by the wedge and stop. The tool is to be placed at right angles to the side of the work, the rubber brought to bear on the face of the rosette. The method of working will be self-evident, after the description already given. It is impossible in a brief work like the present, to go into details of other patterns referred to and illustrated by Bergeron, one or two of which are nevertheless of great beauty, and are executed with the aid of the eccentric chuck mounted on the mandrel of the rose engine. There is, however, a different class of work, to which reference will be made in our next, and we shall also give a description of a simple addition to the slide rest, used by watch case turners, which does away with the necessity of counting the number of divisions upon this instrument when used as above.