Internal or annular gear-wheels have their tooth curves formed by rolling the generating circle upon the pitch circle or base circle, upon the same general principle as external or spur-wheels. But the tooth of the annular wheel corresponds with the space in the spur-wheel, as is shown in Fig. 61, in which curve a forms the flank of a tooth on a spur-wheel p, and the face of a tooth on the annular wheel w. It is obvious then that the generating circle is rolled within the pitch circle for the face of the wheel and without for its flank, or the reverse of the process for spur-wheels. But in the case of internal or annular wheels the path of contact of tooth upon tooth with a pinion having a given number of teeth increases in proportion as the number of teeth in the wheel is diminished, which is also the reverse of what occurs in spur-wheels; as will readily be perceived when it is considered that if in an internal wheel the pinion have as many teeth as the wheel the contact would exist around the whole pitch circles of the wheel and pinion and the two would rotate together without any motion of tooth upon tooth. Obviously then we have, in the case of internal wheels, a consideration as to what is the greatest number (as well as what is the least number) of teeth a pinion may contain to work with a given wheel, whereas in spur-wheels the reverse is again the case, the consideration being how few teeth the wheel may contain to work with a given pinion. Now it is found that although the curves of the teeth in internal wheels and pinions may be rolled according to the principles already laid down for spur-wheels, yet cases may arise in which internal gears will not work under conditions in which spur-wheels would work, because the internal wheels will not engage together. Thus, in Fig. 62, is a pinion of 12 teeth and a wheel of 22 teeth, a generating circle having a diameter equal to the radius of the pinion having been used for all the tooth curves of both wheel and pinion. It will be observed that teeth a, b, and c clearly overlap teeth d, e, and f, and would therefore prevent the wheels from engaging to the requisite depth. This may of course be remedied by taking the faces off the pinion, as in Fig. 63, and thus confining the arc of contact to an arc of recess if the pinion drives, or an arc of approach if the wheel drives; or the number of teeth in the pinion may be reduced, or that in the wheel increased; either of which may be carried out to a degree sufficient to enable the teeth to engage and not interfere one with the other. In Fig. 64 the number of teeth in the pinion p is reduced from 12 to 6, the wheel w having 22 as before, and it will be observed that the teeth engage and properly clear each other.
By the introduction into the figure of a segment of a spur-wheel also having 22 teeth and placed on the other side of the pinion, it is shown that the path of contact is greater, and therefore the angle of action is greater, in internal than in spur gearing. Thus suppose the pinion to drive in the direction of the arrows and the thickened arcs a b will be the arcs of approach, a measuring longer than b. The dotted arcs c d represent the arcs of receding contact and c is found longer than d, the angles of action being 66° for the spur-wheels and 72° for the annular wheel.
On referring again to Fig. 62 it will be observed that it is the faces of the teeth on the two wheels that interfere and will prevent them from engaging, hence it will readily occur to the mind that it is possible to form the curves of the pinion faces correct to work with the faces of the wheel teeth as well as with the flanks; or it is possible to form the wheel faces with curves that will work correctly with the faces, as well as with the flanks of the pinion teeth, which will therefore increase the angle of action, and Professor McCord has shown in an article in the London Engineering how to accomplish this in a simple and yet exceedingly ingenious manner which may be described as follows:—
It is required to find a describing circle that will roll the curves for the flanks of the pinion and the faces of the wheels, and also a describing circle for the flanks of the wheel and the faces of the pinion; the curve for the wheel faces to work correctly with the faces as well as with the flanks of the pinion, and the curve for the pinion faces to work correctly with both the flanks and faces of the internal wheel.
In Fig. 65 let p represent the pitch circle of an annular or internal wheel whose centre is at a, and q the pitch circle of a pinion whose centre is at b, and let r be a describing circle whose centre is at c, and which is to be used to roll all the curves for the teeth. For the flanks of the annular wheel we may roll r within p, while for the faces of the wheel we may roll r outside of p, but in the case of the pinion we cannot roll r within q, because r is larger than q, hence we must find some other rolling circle of less diameter than r, and that can be used in its stead (the radius of r always being greater than the radius of the axis of the wheel and pinion for reasons that will appear presently). Suppose then that in Fig. 66 we have a ring whose bore r corresponds in diameter to the intermediate describing circle r, Fig. 65 and that q represents the pinion. Then we may roll r around and in contact with the pinion q, and a tracing point in r will trace the curve m n o, giving a curve a portion of which may be used for the faces of the pinion. But suppose that instead of rolling the intermediate describing circle r around p, we roll the circle t around p, and it will trace precisely the same curve m n o; hence for the faces of the pinion we have found a rolling circle t which is a perfect substitute for the intermediate circle q, and which it will always be, no matter what the diameters of the pinion and of the intermediate describing circle may be, providing that the diameter of t is equal to the difference between the diameters of the pinion and that of the intermediate describing circle as in the figure. If now we use this describing circle to roll the flanks of the annular wheel as well as the faces of the pinion, these faces and flanks will obviously work correctly together. Since this describing circle is rolled on the outside of the pinion and on the outside of the annular wheel we may distinguish it as the exterior describing circle.
Now instead of rolling the intermediate describing circle r within the annular wheel p for the face curves of the teeth upon p, we may find some other circle that will give the same curve and be small enough to be rolled within the pinion q for its teeth flanks. Thus in Fig. 67 p represents the pitch circle of the annular wheel and r the intermediate circle, and if r be rolled within p, a point on the circumference of r will trace the curve v w. But if we take the circle s, having a diameter equal to the difference between the diameter of r and that of p, and roll it within p, a point in its circumference will trace the same curve v w; hence s is a perfect substitute for r, and a portion of the curve v w may be used for the faces of the teeth on the annular wheel. The circle s being used for the pinion flanks, the wheel faces and pinion flanks will work correctly together, and as the circle s is rolled within the pinion for its flanks and within the wheel for its faces, it may be distinguished as the interior describing circle.
To prove the correctness of the construction it may be noted that with the particular diameter of intermediate describing circle used in Fig. 65, the interior and exterior describing circles are of equal diameters; hence, as the same diameter of describing circle is used for all the faces and flanks of the pair of wheels they will obviously work correctly together, in accordance with the rules laid down for spur gearing. The radius of s in Fig. 69 is equal to the radius of the annular wheel, less the radius of the intermediate circle, or the radius from a to c. The radius of the exterior describing circle t is the radius of the intermediate circle less the radius of the pinion, or radius c b in the figure.
Now the diameter of the intermediate circle may be determined at will, but cannot exceed that of the annular wheel or be less than the pinion. But having been selected between these two limits the interior and exterior describing circles derived from it give teeth that not only engage properly and avoid the interference shown in Fig. 62, but that will also have an additional arc of action during the recess, as is shown in Fig. 68, which represents the wheel and pinion shown in Fig. 62, but produced by means of the interior and exterior describing circles. Supposing the pinion to be the driver the arc of approach will be along the thickened arc of the interior describing circle, while during the arc of recess there will be an arc of contact along the dotted portion of the exterior describing circle as in ordinary gearing. But in addition there will be an arc of recess along the dotted portion of the intermediate circle r, which arc is due to the faces of the pinion acting upon the faces as well as upon the flanks of the wheel teeth. It is obvious from this that as soon as a tooth passes the line of centres it will, during a certain period, have two points of contact, one on the arc of the exterior describing circle, and another along the arc of r, this period continuing until the addendum circle of the pinion crosses the dotted arc of the exterior describing circle at z.
The diameters of the interior and exterior describing circles obviously depend upon the diameter of the intermediate circle, and as this may, as already stated, be selected, within certain limits, at will, it is evident that the relative diameters of the interior and exterior describing circles will vary in proportion, the interior becoming smaller and the exterior larger, while from the very mode of construction the radius of the two will equal that of the axes of the wheel and pinion. Thus in Fig. 69 the radii of s, t, equal a b, or the line of centres, and their diameters, therefore, equal the radius of the annular wheel, as is shown by dotting them in at the upper half of the figure. But after their diameters have been determined by this construction either of them may be decreased in diameter and the teeth of the wheels will clear (and not interfere as in Fig. 62), but the action will be the same as in ordinary gear, or in other words there will be no arc of action on the circle r. But s cannot be increased without correspondingly decreasing t, nor can t be increased without correspondingly decreasing s.
Fig. 70 shows the same pair of gears as in Fig. 68 (the wheel having 22 and the pinion 12 teeth), the diameter of the intermediate circle having been enlarged to decrease the diameter of s and increase that of t, and as these are left of the diameter derived from the construction there is receding action along r from the line of centres to t.
In Fig. 71 are represented a wheel and pinion, the pinion having but four teeth less than the wheel, and a tooth, j, being shown in position in which it has contact at two places. Thus at k it is in contact with the flank of a tooth on the annular wheel, while at l it is in contact with the face of the same tooth.
As the faces of the teeth on the wheel do not have contact higher than point t, it is obvious that instead of having them 3⁄10 of the pitch as at the bottom of the figure, we may cut off the portion x without diminishing the arc of contact, leaving them formed as at the top of the figure. These faces being thus reduced in height we may correspondingly reduce the depth of flank on the pinion by filling in the portion g, leaving the teeth formed as at the top of the pinion. The teeth faces of the wheel being thus reduced we may, by using a sufficiently large intermediate circle, obtain interior and exterior describing circles that will form teeth that will permit of the pinion having but one tooth less than the wheel, or that will form a wheel having but one tooth more than the pinion.
The limits to the diameter of the intermediate describing circle are as follows: in Fig. 72 it is made equal in diameter to the pitch diameter of the pinion, hence b will represent the centre of the intermediate circle as well as of the pinion, and the pitch circle of the pinion will also represent the intermediate circle r. To obtain the radius for the interior describing circle we subtract the radius of the intermediate circle from the radius of the annular wheel, which gives a p, hence the pitch circle of the pinion also represents the interior circle r. But when we come to obtain the radius for the exterior describing circle (t), by subtracting the radius of the pinion from that of the intermediate circle, we find that the two being equal give o for the radius of (t), hence there could be no flanks on the pinion.
Now suppose that the intermediate circle be made equal in diameter to the pitch circle of the annular wheel, and we may obtain the radius for the exterior describing circle t; by subtracting the radius of the pinion from that of the intermediate circle, we shall obtain the radius a b; hence the radius of (t) will equal that of the pinion. But when we come to obtain the radius for the interior describing circle by subtracting the radius of the intermediate circle from that of the annular wheel, we find these two to be equal, hence there would be no interior describing circle, and, therefore, no faces to the pinion.
The action of the teeth in internal wheels is less a sliding and more a rolling one than that in any other form of toothed gearing. This may be shown as follows: In Fig. 73 let a a represent the pitch circle of an external pinion, and b b that of an internal one, and p p the pitch circle of an external wheel for a a or an internal one for b b, the point of contact at the line of centres being at c, and the direction of rotation p p being as denoted by the arrow; the two pinions being driven, we suppose a point at c, on the pitch circle p p, to be coincident with a point on each of the two pinions at the line of centres. If p p be rotated so as to bring this point to the position denoted by d, the point on the external pinion having moved to e, while that on the internal pinion has moved to f, both having moved through an arc equal to c d, then the distance from e to d being greater than from d to f, more sliding motion must have accompanied the contact of the teeth at the point e than at the point f; and the difference in the length of the arc e d and that of f d, may be taken to represent the excess of sliding action for the teeth on e; for whatever, under any given condition, the amount of sliding contact may be, it will be in the proportion of the length of e d to that of f d. Presuming, then, that the amount of power transmitted be equal for the two pinions, and the friction of all other things being equal—being in proportion to the space passed (or in this case slid) over—it is obvious that the internal pinion has the least friction.
In Fig. 74 are shown a worm and worm gear partly in section on the line of centres. The worm or tangent screw w is simply one long tooth wound around a cylinder, and its form may be determined by the rules laid down for a rack and pinion, the tangent screw or worm being considered as a rack and the wheel as an ordinary spur-wheel.
Worm gearing is employed for transmitting motion at a right angle, while greatly reducing the motion. Thus one rotation of the screw will rotate the wheel to the amount of the pitch of its teeth only. Worm gearing possesses the qualification that, unless of very coarse pitch, the worm locks the wheel in any position in which the two may come to a state of rest, while at the same time the excess of movement of the worm over that of the wheel enables the movement of the latter, through a very minute portion of a revolution. And it is evident that, when the plane of rotation of the worm is at a right angle to that of the wheel, the contact of the teeth is wholly a sliding one. The wear of the worm is greater than that of the wheel, because its teeth are in continuous contact, whereas the wheel teeth are in contact only when passing through the angle of action. It may be noted, however, that each tooth upon the worm is longer than the teeth on the wheel in proportion as the circumference of the worm is to the length of wheel tooth.
If the teeth of the wheel are straight and are set at an angle equal to the angle of the worm thread to its axis, as in Fig. 75, p p representing the pitch line of the worm, c d the line of centres, and d the worm axis, the contact of tooth upon tooth will be at the centre only of the sides of the wheel teeth. It is generally preferred, however, to have the wheel teeth curved to envelop a part of the circumference of the worm, and thus increase the line of contact of tooth upon tooth, and thereby provide more ample wearing surface.
In this case the form of the teeth upon the worm wheel varies at every point in its length as the line of centres is departed from. Thus in Fig. 76 is shown an end view of a worm and a worm gear in section, c d being the line of centres, and it will be readily perceived that the shape of the teeth if taken on the line e f, will differ from that on the line of centres c d; hence the form of the wheel teeth must, if contact is to occur along the full length of the tooth, be conformed to fit to the worm, which may be done by taking a series of section of the worm thread at varying distances from, and parallel to, the line of centres and joining the wheel teeth to the shape so obtained. But if the teeth of the wheel are to be cut to shape, then obviously a worm may be provided with teeth (by serrating it along its length) and mounted in position upon the wheel so as to cut the teeth of the wheel to shape as the worm rotates. The pitch line of the wheel teeth, whether they be straight and are disposed at an angle as in Fig. 75, or curved as in Fig. 76, is at a right angle to the line of centres c d, or in other words in the plane of g h, in Fig. 76. This is evident because the pitch line must be parallel to the wheel axis, being at an equal radius from that axis, and therefore having an equal velocity of rotation at every point in the length of the pitch line of the wheel tooth.
If we multiply the number of teeth by their pitch to obtain the circumference of the pitch circle we shall obtain the circumference due to the radius of g h, from the wheel axis, and so long as g h is parallel to the wheel axis we shall by this means obtain the same diameter of pitch circle, so long as we measure it on a line parallel to the line of centres c d. The pitch of the worm is the same at whatever point in the tooth depth it may be measured, because the teeth curves are parallel one to the other, thus in Fig. 77 the pitch measures are equal at m, n, or o.
But the action of the worm and wheel will nevertheless not be correct unless the pitch line from which the curves were rolled coincides with the pitch line of the wheel on the line of centres, for although, if the pitch lines do not so coincide, the worm will at each revolution move the pitch line of the wheel through a distance equal to the pitch of the worm, yet the motion of the wheel will not be uniform because, supposing the two pitch lines not to meet, the faces of the pinion teeth will act against those of the wheel, as shown in Fig. 78, instead of against their flanks, and as the faces are not formed to work correctly together the motion will be irregular.
The diameter of the worm is usually made equal to four times the pitch of the teeth, and if the teeth are curved as in figure 76 they are made to envelop not more than 30° of the worm.
The number of teeth in the wheel should not be less than thirty, a double worm being employed when a quicker ratio of wheel to worm motion is required.
When the teeth of the wheel are curved to partly envelop the worm circumference it has been found, from experiments made by Robert Briggs, that the worm and the wheel will be more durable, and will work with greatly diminished friction, if the pitch line of the worm be located to increase the length of face and diminish that of the flank, which will decrease the length of face and increase the length of flank on the wheel, as is shown in Fig. 79; the location for the pitch line of the worm being determined as follows:—
The full radius of the worm is made equal to twice the pitch of its teeth, and the total depth of its teeth is made equal to .65 of its pitch. The pitch line is then drawn at a radius of 1.606 of the pitch from the worm axis. The pitch line is thus determined in Fig. 76, with the result that the area of tooth face and of worm surface is equalized on the two sides of the pitch line in the figure. In addition to this, however, it may be observed that by thus locating the pitch line the arcs both of approach and of recess are altered. Thus in Fig. 80 is represented the same worm and wheel as in Fig. 79, but the pitch lines are here laid down as in ordinary gearing. In the two figures the arcs of approach are marked by the thickened part of the generating circle, while the arcs of recess are denoted by the dotted arc on the generating circle, and it is shown that increasing the worm face, as in Fig. 79, increases the arc of recess, while diminishing the worm flank diminishes the arc of approach, and the action of the worm is smoother because the worm exerts more pulling than pushing action, it being noted that the action of the worm on the wheel is a pushing one before reaching, and a pulling one after passing, the line of centres.
It may here be shown that a worm-wheel may be made to work correctly with a square thread. Suppose, for example, that the diameter of the generating circle be supposed to be infinite, and the sides of the thread may be accepted as rolled by the circle. On the wheel we roll a straight line, which gives a cycloidal curve suitable to work with the square thread. But the action will be confined to the points of the teeth, as is shown in Fig. 81, and also to the arc of approach. This is the same thing as taking the faces off the worm and filling in the flanks of the wheel. Obviously, then, we may reverse the process and give the worm faces only, and the wheel, flanks only, using such size of generating circle as will make the spaces of the wheel parallel in their depths and rolling the same generating circle upon the pitch line of the worm to obtain its face curve. This would enable the teeth on the wheel to be cut by a square-threaded tap, and would confine the contact of tooth upon tooth to the recess.
The diameter of generating circle used to roll the curves for a worm and worm-wheel should in all cases be larger than the radius of the worm-wheel, so that the flanks of the wheel teeth may be at least as thick at the root as they are at the pitch circle.
To find the diameter of a wheel, driven by a tangent-screw, which is required to make one revolution for a given number of turns of the screw, it is obvious, in the first place, that when the screw is single-threaded, the number of teeth in the wheel must be equal to the number of turns of the screw. Consequently, the pitch being also given, the radius of the wheel will be found by multiplying the pitch by the number of turns of the screw during one turn of the wheel, and dividing the product by 6.28.
When a wheel pattern is to be made, the first consideration is the determination of the diameter to suit the required speed; the next is the pitch which the teeth ought to have, so that the wheel may be in accordance with the power which it is intended to transmit; the next, the number of the teeth in relation to the pitch and diameter; and, lastly, the proportions of the teeth, the clearance, length, and breadth.
When the amount of power to be transmitted is sufficient to cause excessive wear, or when the velocity is so great as to cause rapid wear, the worm instead of being made parallel in diameter from end to end, is sometimes given a curvature equal to that of the worm-wheel, as is shown in Fig. 82.
The object of this design is to increase the bearing area, and thus, by causing the power transmitted to be spread over a larger area of contact, to diminish the wear. A mechanical means of cutting a worm to the required form for this arrangement is shown in Fig. 83, which is extracted from “Willis’ Principles of Mechanism.” “a is a wheel driven by an endless screw or worm-wheel, b, c is a toothed wheel fixed to the axis of the endless screw b and in gear with another and equal toothed gear d, upon whose axis is mounted the smooth surfaced solid e, which it is desired to cut into Hindley’s[2] endless screw. For this purpose a cutting tooth f is clamped to the face of the wheel a. When the handle attached to the axis of b c is turned round, the wheel a and solid wheel e will revolve with the same relative velocity as a and b, and the tool f will trace upon the surface of the solid e a thread which will correspond to the conditions. For from the very mode of its formation the section of every thread through the axis will point to the centre of the wheel a. The axis of e lies considerably higher than that of b to enable the solid e to clear the wheel a.
[2] The inventor of this form of endless screw.
“The edges of the section of the solid e along its horizontal centre line exactly fit the segment of the toothed wheel, but if a section be made by a plane parallel to this the teeth will no longer be equally divided as they are in the common screw, and therefore this kind of screw can only be in contact with each tooth along a line corresponding to its middle section. So that the advantage of this form over the common one is not so great as appears at first sight.
“If the inclination of the thread of a screw be very great, one or more intermediate threads may be added, as in Fig. 84, in which case the screw is said to be double or triple according to the number of separate spiral threads that are so placed upon its surface. As every one of these will pass its own wheel-tooth across the line of centres in each revolution of the screw, it follows that as many teeth of the wheel will pass that line during one revolution of the screw as there are threads to the screw. If we suppose the number of these threads to be considerable, for example, equal to those of the wheel teeth, then the screw and wheel may be made exactly alike, as in Fig. 85; which may serve as an example of the disguised forms which some common arrangements may assume.”
In Fig. 86 is shown Hawkins’s worm gearing. The object of this ingenious mechanical device is to transmit motion by means of screw or worm gearing, either by a screw in which the threads are of equal diameter throughout its length, or by a spiral worm, in which the threads are not of equal diameter throughout, but increase in diameter each way from the centre of its length, or about the centre of its length outwardly. Parallel screws are most applicable to this device when rectilinear motions are produced from circular motions of the driver, and spiral worms are applied when a circular motion is given by the driver, and imparted to the driven wheel. The threads of a spiral worm instead of gearing into teeth like those of an ordinary worm-wheel, actuate a series of rollers turning upon studs, which studs are attached to a wheel whose axis is not parallel to that of the worm, but placed at a suitable inclination thereto. When motion is given to the worm then rotation is produced in the roller wheel at a rate proportionable to the pitch of worm and diameter of wheel respectively.
In the arrangement for transmitting rectilinear motion from a screw, rollers may be employed whose axes are inclined to the axis of the driving screw, or else at right angles to or parallel to the same. When separate rollers are employed with inclined axes, or axes at right angles with that of the main driving screw, each thread in gear touches a roller at one part only; but when the rollers are employed with axes parallel to that of the driving screw a succession of grooves are turned in these rollers, into which the threads of the driving screw will be in gear throughout the entire length of the roller. These grooves may be separate and apart from each other, or else form a screw whose pitch is equal to that of the driving screw or some multiple thereof.
In Fig. 86 the spiral worm is made of such a length that the edge of one roller does not cease contact until the edge of the next comes into contact; a wheel carries four rollers which turn on studs, the latter being secured by cottars; the axis of the worm is at right angles with that of the wheel. The edges of the rollers come near together, leaving sufficient space for the thread of the worm to fit between any two contiguous rollers. The pitch line of the screw thread forms an arc of a circle, whose centre coincides with that of the wheel, therefore the thread will always bear fairly against the rollers and maintain rolling contact therewith during the whole of the time each roller is in gear, and by turning the screw in either direction the wheel will rotate.
To prevent end thrust on a worm shaft it may have a right-hand worm a, and a left-hand one c (Fig. 87), driving two wheels b and d which are in gear, and either of which may transmit the power. The thrust of the two worms a and c, being in opposite directions, one neutralizes the other, and it is obvious that as each revolution of the worm shaft moves both wheels to an amount equal to the pitch of the worms, the two wheels b d may, if desirable, be of different diameters.
Involute teeth.—These are teeth having their whole operative surfaces formed of one continuous involute curve. The diameter of the generating circle being supposed as infinite, then a portion of its circumference may be represented by a straight line, such as a in Fig. 88, and if this straight line be made to roll upon the circumference of a circle, as shown, then the curve traced will be involute p. In practice, a piece of flat spring steel, such as a piece of clock spring, is used for tracing involutes. It may be of any length, but at one end it should be filed so as to leave a scribing point that will come close to the base circle or line, and have a short handle, as shown in Fig. 89, in which s represents the piece of spring, having the point p′, and the handle h. The operation is, to make a template for the base circle, rest this template on drawing paper and mark a circle round its edge to represent on the paper the pitch circle, and to then bend the spring around the circle b, holding the point p′ in contact with the drawing paper, securing the other end of the piece of steel, so that it cannot slip upon b, and allowing the steel to unwind from the cylinder or circle b. The point p′ will mark the involute curve p. Another way to mark an involute is to use a piece of twine in place of the spring and a pencil instead of the tracing point; but this is not so accurate, unless, indeed, a piece of wood be laid on the drawing-board and the pencil held firmly against it, so as to steady the pencil point and prevent the variation in the curve that would arise from variation in the vertical position of the pencil.
The flanks being composed of the same curve as the faces of the teeth, it is obvious that the circle from which the tracing point starts, or around which the straight line rolls, must be of less diameter than the pitch circle, or the teeth would have no flanks.
A circle of less diameter than the pitch circle of the wheel is, therefore, introduced, wherefrom to produce the involute curves forming the full side of the tooth.
The depth below pitch line or the length of flank is, therefore, the distance between the pitch circle and the base circle. Now even supposing a straight line to be a portion of the circumference of a circle of infinite diameter or radius, the conditions would here appear to be imperfect, because the generating circle is not rolled upon the pitch circle but upon a circle of lesser diameter. But it can be shown that the requirements of a proper velocity ratio will be met, notwithstanding the employment of the base instead of the pitch circle. Thus, in Fig. 90, let a and b represent the respective centres of the two pitch circles, marked in dotted lines. Draw the base circle for b as e q, which may be of any radius less than that of the pitch circle of b. Draw the straight line q d r touching this base circle at its perimeter and passing through the point of contact on the pitch circles as at d. Draw the circle whose radius is a r forming the base circle for wheel a. Thus the line r p q will meet the perimeters of the two circles while passing through the point of contact d at the line of centres (a condition which the relative diameters of the base circles must always be so proportioned as to attain).
If now we take any point on r q, as p in the figure, as a tracing point, and suppose the radius or distance p q to represent the steel spring shown in Fig. 89, and move the tracing point back to the base circle of b, it will trace the involute e p. Again we may take the tracing point p (supposing the line p r to represent the steel spring), and trace the involute p f, and these two involutes represent each one side of the teeth on the respective wheels.