US2011956A - Method of grinding worms and screws - Google Patents

Description

Aug. 20, 1935, N. TRBOJIEVICH 2,011,956

METHOD OF GRINDING WORMS AND SCREWS Filed July 7, 1930 s Sheets-Sheet 1 w F W ATTORNEY/8' Aug. 20, 1935. N. TRBOJEVICH METHOD or GRINDING WORMS AND SCREWS Filed July 7, 1930 s Sheets-Sheet 2 INVENTOR ATTORNEYS Aug. 20, 1935. N. TRBOJEVICH METHOD OF GRINDING WORMS'AND SCREWS 3 Sheets-Sheet 3 Filed July '7, 1950 zum mm t I ATTORNEYS Patented Aug. 20, 1935 UNITED-"STATES METHOD OF GRINDING WORMS SCREWS Nikola Trbojevich, Detroit, Mich. Application July '2, 1930, Serial No. 466,204

'1 Claims.

The invention relates to a novel method of grinding worms and screws.

Heretofore such worms were ground by means I, of a disk wheel which was positioned in its plane a at an angle relative to theworm axis and the work was rotatedand translated relative to the wheel in a succession of rapid passes. each pass the machine would stop and the work (or the wheel) would be returned to its initial position and another pass would be taken. In order to prevent the burning or cracking of the 'work the passes had tobe taken at a considerable rate of speed, from 15 to '70 ft. per minute,

thus giving passes of a very short duration, especially so in the worms that were comparatively short and had a steep helix angle, as in automobile axle worms, for instance.

I have found three fundamental disadvantages existing in the above described conventional method and have succeeded in eliminating all three of them by improving the method of grinding as hereinafter shown. The first disadvantage, the short duration of individual operasecond disadvantage is that the form as ground in the worm thread depends not only upon the exact shape to which the wheel is dressed but also upon the diameter of the wheel, the so-called helical interference. This isa considerable source .of trouble because the wheel is reduced in diameter each time it is dressed. thus producing slightly different workeach time. The third disadvantage is that the wheel engages the work i with a line contact thus preventing the ground ofl particles from leaving the place of contact instantaneously and causing overheating, clogging and scratching.

In my improved method I position the wheel in the same fashion as before, but I oscillate the same in its plane tangentially and transversely first change the nature of wheel contact completely by going from a line to point contact; second, the produced form is now independent of the wheel diameter and third, the duration of each pass may now be made as long as desired and the entire stock may be removed in one pass only After I tions or passes, I have already mentioned. The

work, to save labor and wear and tear upon the 5 machine and to reduce the expense for emery wheels.

In the drawings I I Figure 1 is the transverse section of the worm taken in the plane ll of Figure 2 10" Figure 2 is the side view of the worm diagrammatically represented;

Figure 3 is the cross section of a grinder capable of generating the worm shown in Figures 1 and 2; 5

Figure 4 is a diagram showingthegrinder and the imaginary prism mating with the worm in plane 4-4 of Figure 2;

' Figures 5 and 6 ared'iagrams explanatory oi? the Equations 4 and 5; 201' Figure 7 is a perspective view of an involute helicoid;

Figure 8 is the plan view of my improved worm grinder;

Figure 9 is the section 9-4 of Figure 8 showing 25 the feed mechanism employed in this machine;

Figure 10 is a diagrammatic view of a globoid A v worm being ground on this principle.

The principle upon which this invention is based will now-be explained. In Figure 1 the trans- 3o bounded on its two sides by two similar trans- 35 verse tooth curves 23 which are preferably involutes developed from the base circle 24, but they also may be any other curves such' as Archimedean and other spirals without afiecting this broad principle. A pitch circle 25 is now selected or established and the tooth curves 23 are fixed in their relative positions by assuming a thickness of tooth u measured along the arc of the said pitch circle, while the active or contact surface of the said curves will extend from the outside circle 28 to within a short distance from the root circle 21, thus providing a clearance space at the bottoms of the worm teeth.

In order to generate a worm from the said i lamina 2| we rotate and translate the same at a fixed ratio about the axis 3i perpendicular to and concentric with the lamina, thus producing an infinite number of helixes 30 all having the same lead L, Figure 2. The lamina2l may be considered as a spur gear element and thus it will correctly mesh with a rack 28 having a pitch line 29 tangent to the pitch circle at the point C, a pressure angle A equal to the pressure angle of the tooth curve 23 at the point B where the said curve intersects the pitch circle 25, and a tooth flank 32 conjugate to the said curve 23. The flank 32 will be a. straight line .if the curve 23 is an involute and will be a curve in all other cases.

It is now evident that inasmuch as the lamina 2| is rotated and propagated uniformly, the rack 28 in order to remain in mesh with the said lamina wfll have to be moved in two mutually perpendicular directions, i. e. along its pitch line 29 and the worm axis 3|. Thus, the rack 28 will describe a prism having an axis coincident with the line 33,-Figure 2, and an angle of inclination D corresponding to the helical angle of the worm. If we now denote the width of space BE of the lamina with t, measured along the arc the thickness of the rack tooth BE will also be equal to t and the pitch of the rack p=B'F will be equal to where r is the pitch radius and n the number of teeth or lobes in the lamina.

It-stands further that 8 cos A- I (2) where a is the base radius, and

L tan D- m (3) To determine the cross section of the grinder 34, Figure 3 we must know its thickness-t and its pressure angle A. From the triangles Figures and 6, we have tan A=tan A sin D t'=t Sin D and G'C"=GC, Fig. 1

Now, the object of this invention is, first, to find the prism meshing with the worm to be ground, second, to mechanically reproduce the said prism by means of an oscillating grinder, third, to determine the necessary amplitude or length of stroke of the said grinder and fourth, to rotate and translate the worm in order to finish both sides of the thread from end to end in one cut.

The next step will be to determine the lines of contact of the prism 35, Figure 4 with the worm threads. In order to do this we begin with the Figure 1 in which the two lines of action HCM and KCN are first determined. These two lines will be both straight and tangent to the base circle in the involute system, will be certain curves in any other system and will cross each other at the pitch point C in all systems.

When the lamina 2| is rotated in the direction of the arrow 36 the rack will translate in unison in the direction of the arrow 31 and the points of tangency will move along the two pressure lines HCM and KCN, thus requiring the projected length of stroke S to completely finish the curves 23 from top to bottom. Therefore, if we project the said pressure lines upon the worm thread surface in Figure 2 we obtain the two skew lines NK' and HM' respectively, said lines being the locus of the simultaneous contact of the worm thread with the prism 35 at anyone instant.

Although the thread surfaces 38 and 39 are warped and curved it is readily proved that in involute system both lines of contact NK and HM' respectively are straight lines and are oppositely inclined relative to the plane of paper, Figure 2 without intersecting each other. That these lines are straight follows from Figure 7 which diagrammatically represents another well known method of generating aninvolute helicoid. A right angle triangle PP'P rolls upon the base cylinder 40 in such a manner that its adjacent side PP rolls upon the base circle 24 and its hypothenuse P'P" rolls upon the base helix 4| thereby describing with the apex P an involute 23 in the plane of the said base circle. The helicoidal surface is, therefore, composed of straight lines PP" which are at each instant perpendicular to the tangent 32 to the involute 23, said tangent being identical with the rack tooth flank 32 as shown in Figure 1. Due to the fact that the developed length of the generator FF is exactly equal to the arc length of the helix 4| it follows from the theory of such surfaces (developables) that the plane comprising the perpendicular lines 32 and PT" is tangent to the surface and the line of tangency is along the generator PP".

It will be of interest to note how I increase the productive ability of this type of the machine up to its theoretical maximum. This maximum will be attained when the projected length of the stroke S, Figure 1 will be the minimum. Therefore, I select the pitch circle 25 in such a relation to the depth of thread, that the projections of the points HNKM will produce the minimum amplitude as measured in the direction of the rack pitch line 29. This will happen when the pitch line is located at a certain point not far removed from the middle of the working depth.

To summarize the results of this investigation in the theory, we are now enabled to select the pitch line properly i. e. such that the length of stroke will be the minimum and the two opposite sides of the thread will be simultaneously generated; we also may compute the length of the stroke S, Figure 4 the normal thickness 1." and the normal pressure angle A of the wheel.

Figure 8 shows the plan view of a simple and yet operative machine for grinding worms in accordance with my invention. This machine may be constructed from an ordinarysurface grinding machine and it will be found to be reasonably accurate and automatic (during one complete pass) in its operation.

The work table 42 of the machine reciprocates in a direction at right angles to the wheel axis 43 in the bed 440: of the machine and the number, speed and length of strokes are adjusted in the same manner as in any other surface grinder. The grinding wheel 34 may be of any desired diameter and should be rapidly rotated to have a peripheral speed of about five to six thousand feet per minute in order to grind hardened steel. The wheel contour is V-shaped and is dressed to the exact dimensions as was shown in Figure 3. The worm to be ground 44 is mounted upon the arbor 45. It abuts the shoulder 46 with its one end and is kept tightly clamped in position by means of a nut 41 at its other end. The rear end of the arbor 45 is formed in a mandrel 48 and carries a master screw 49 which is tightly clamped upon the said mandrel by means of the nut 50. The diameter, the form of thread and the number of threads in-the master 49 may be and feed the worm the above mentioned four I arbitrarily'selected; however, it is necessary that its lead be exactly the same as that of the worm 44 and its d also. must be the. same.

The master snugly fits into the nut which is preferably made of cast iron lined with babbitt and is clamped by means of screw 52 in a bore situated in the upper part of the vertical shank 53 of the base plate 54.

The front end of the arbor 45 is formed into a slidingspindle'55 and, is provided with one or more longitudinal keys 56.- The feed worm gear 51 is of the self-locking (single thread) type, it'

is keyed to the spindle by means or the key I8 (Figure9), and-is rotatably housed in the corresponding bore of the bracket 59, beingheld therein position, by meansof the thrust collar II; The feed screw GHFigure 9) is formedin a "taperfiournal .62 at its upper end where it rotatabiy fits in the correspondingtaper bearing of .thebrack'et 59, an is held in position by means of the. thrust 00 r ,63. The said screw 6| snugly fits into'the teeth of the worm gear- 51 with little or ,no backlash andis pivoted by means of the'pivot 64 in the base 65 of the s bracket 59. A ratche t 66 is keyed by means of a key 61 to the said i'eed screw and'serves to operate the same at suitable intervals. The said screw is'also operable by hand by means of the squareshank .88; formed at its upper end. The

' bracket 59 is securely bolted'to the base plate 54 by means of. bolts 69.

The operation of this machine will now be ex plained'. The axis\of the worm 44 is inclined at the helix angle D relative to the axis 43 of the wheel and theworm. is reciprocated in the directions 6f the arrows I0. and II, perpendicular tothe said wheelaxis. Thus, the wheel 34 describes a prism ora rack tooth tangential to the wo thread'in the space. Upon each stroke of the table 42. two'lines N'K' and HM respectively (see Figure 2) will be ground in the worm thread and it we now slowly-rotate and trans-. late the worm along its axis, upon the next stroke another two lines will be generated adjacent to the above mentioned lines NK' and'H'M' and also lying in the worm thread.

In the practical example shown in Figure 8, I have 100 teeth in the feed worm gear'and 16 teeth in the ratchet, thus obtaining a ratio of 1600 to 1 giving a spacing of generated lines in the finished worm thread about four thousandths of I an inch apart for a length of helix of 6 inches.

the screw ii, the master screw 49 will rotatm andalso advance along its axis in the nut 5|, thus carrying the worm to be ground 44 in the required and precise helical path. The ratchet 66 will engage the adjustable stop I2 once during each alternate stroke in the direction of. the arrow in this design, .thus causing the ratchet 66 to ro-- tate through an angle corresponding toone tooth thousandths of an inch.

The ratchet feed mechanism shown in Figures 8 and 9 is only a modification or a design and it will be understood that this apparatus will properly work also in the case when the worm 44 is rotated continuously, e. g. by a train of gears from an outside source instead of. periodically as by the ratchet. It will also be understood that instead of reciprocating the worm in the direction of the arrows"! and 'Ii we may reciproc te the wheel 34 in the same direction without afiectin the principle. The contours 13 ofthe grinder will be straight lines to generate worms which are involute in their transverse section, but will be curved for all other types of worms as already mentioned. In ordinary work, the worm is first roughed out in soft, hardened, and ground only in order to remove a very thin film-of metal, usually less than about ten thousandths of an inch thick; however, in grinding high speed steel taps iromthe solid steel, the whole depth of thread may be taken in one cut in this method. This ability of removingzcemparatively large amounts of SlJOCkIlIl one pass, I consider asthe greatest advantage of this method. and this advantage is not shared by any other-known method of worm or screw grind- Globoid worms [may also be ground by this method. As isdiagrammatically shown in Figure '10 the globoidjworm is'of such a form that it will contact with a prism along two oppositely siti. e. it is translatedin'aigloboid helix instead of in a cylindrical or ordinary helix as for common Thus both sides of the wormthread may be" ground in one pass from end to end.

When the worm, straight or globoid, is multiple threaded, each thread isground in a separate operation providing one grinding wheel only is used. In Figure 8 the worm 44 is double threaded and after grinding one thread we disengage the nut 41 and index the work in order to be able to grind the other thread without disturbing the wheel 34 in its position.

What I claim as my invention is:

1. A method of grinding worms which consists in forming a disk grinder. to a profile capable of touching an imaginary prism element along a line, said prism in turn being capable of touching the helicalsurface to be ground along another line, in positioning thecutting plane of the grinder to coincide with the axis of the said prism, in oscillating the said grinder along the said prism axis and in the plane of the grinder in a direction transverse and tangential relative to the root cylinder of they work thereby reproducing the prism and in translating the worm along its axis in a helix untilthe thread is finished from end to end in the form of a series of generated lines.

2. A method of grinding a helical thread consisting in forming a disk grinder in its axial plane to a profile coinciding with theinormal section of an imaginary prism member which member is capable of. touching the worm thread surface along two-lines, one line on eachside of the thread, in reciprocating. the said grinder parallel to the axis of the said prism and tangentially of fthe root cylinder of the work in a stroke exceeding in length the projected theoretical length of contact and in imparting to the worm a slow helical translation along the axis, thereby finishing two worm thread surfaces from end to end in the form of a series of generated lines.

3. A'method of grinding worms, screws and the like in which the grinding wheel is positioned in a direction tangentiar of the worm thread and transverse relative to the axis of worm, in which the worm is slowly translated relative to the wheelin a predetermined helical path, in which the wheel is oscillated in its plane and'in a direction-tangential of the root cylinder of the work, in which the rate of oscil lation of the grinder is of a sufficient rapidity to provide an adequate surface speed relative to the work to prevent burning and in which the length of the stroke is selected to be in excess of the length of engagement of the worm thread with an imaginary prism member which the oscillating grinder represents, thereby making it possible to remove by grinding comparatively large amounts of stock in one passage of the worm in its helical path past the grinder.

4. A method of generating helical thread surfaces having a plurality of thread convolutions disposed about an axis in which a disk cutter is formed to correspond in form to the normal section of a rack tooth capable of meshing with the said worm, in which the cutter is reciprocated in the direction of the rack tooth axis tangentially of the root cylinder of the work and transversely of the said worm axis and in which the worm is slowly translated in a helical path past the cutter to finish the thread from end to end in a series of generated lines, thereby eliminating the efiects of the helical interference and enabling the operator to produce an identical tooth form in the worm by using a cutter of a predetermined profile but entirely disregarding the diameter of the said cutter.

5. A method of grinding worms whose threads are wound along a circular helix and have involute cross contours as measured in the plane perpendicular to the worm axis in which a disk grinder is formed to a double conical shape, each cone angle corresponding to the normal pressure angle of the worm to be ground, and having a thickness of disk as measured over the two cones corresponding to the normal width of space of the said worm; in which the grinder is positioned tangentially of the worm thread at an angle equal to the helix angle of the worm,.

- prising imparting a helicoidal translation to the work along its axis and rapidly oscillating the grinder in its plane of rotation in a direction tangent to the root cylinder of the work relative to the work.

7. A method of grinding worms which consists in slowly feeding the work relative to a rotary grinder in a helical path along thev axis of the work and in relatively rapidly oscillating the grinder in its plane of rotation in a straight line path in a transverse direction to the work axis and tangent to the root cylinder of the work, thus constantly shifting the point of momentary engagement whereby the heat is effectively dissipated from the work and grinder.

NIKOLA TRBOJEVICH.

Images