ELEMENTS OF SPHERICAL TURNING - Model Engineer

ELEMENTS OF SPHERICAL TURNINGA review of methods and appliances for

machining contours involving circular arcs

by Edgar T. Westbury

T HE NEED to machine components to sphericalcontours is often encountered in mechanical engi-neering, and many devices varying in design andcomplexity, have been produced for this purpose.Several of these have been described at varioustimes in Model Engineer, but in view of the manyqueries concerning spherical turning which havearisen, there is a call for a general review of theprinciples employed and the methods applicableto particular kinds of work.

to finished shape as possible, and then chuckedcrosswise, and at other angles, t o its original axis,for the correction of spherical errors. Theseoperations, needless to say, call for a special skillof both hand and eye, they may, perhaps, havenow been completely superseded by modern pro-duction processes, but were certainly employed upto quite recent years.

Cutting angles

Freehand methodsA great deal obviously depends on the degree

of accuracy required in the spherical contour ofthe work, and when using simple devices, this callsfor considerable skill on the part of the operator.More positive accuracy can be obtained by deviceswith mechanical control of t h e tool movement,but only provided that these are well designed andproperly adjusted. In t h i s respect, the resultsobtained are not always of a quality commensur-ate with the elaboration of the apparatus em-ployed. Simple hand-controlled tools will oftenproduce a standard of accuracy sufficient fo rpractical purposes. In wood turning, for instance,the appearance is the main, or possibly the only,thing that matters, and the skilled turner obtainsthe desired result with no other tools than theroughing-out gouge and the oblique-edged chisel(Fig. 1). These are normally applied at a tangentnear the top surface of the work and tilted orswivelled on the hand rest to produce convexcontours.

The tools and techniques employed in turningwood and other relatively soft materials (Fig. 2A)are not so well suited to working on hard metals,in which cutting angles, and to some extentmethods of application, are necessarily different.Tools for metal are usually applied with the top

PRONG DRIVING CENTRE

TURNING CHISE

Most spherical contours comprise only a partof a complete sphere, with a supporting stem orstalk. Such parts as balanced handles for machinetools, stanchions, o r handrai1 knobs, and ball-ended rods and links are in this category. Theproduction of a complete ball is not often neces-sary in mechanical turning practice, because ballsof precise accuracy, in a w i d e variety of sizes andvarious materials, can generally be obtained readymade. It may, however, be worth while to referbriefly to the production of such items as billiardballs, in which high accuracy has been obtainedby the use of hand tools alone. Cup chucks, oftenmade in hardwood, are generally employed tohold t h e material, which is roughed out as closely

Above: Fig. I. Below: Fig. 2

A

MODEL ENGINEER 5 May 1967 425

cutting edge near level with the horizontal centre-line of the work (Fig. 2B), and the cutting anglesare much more obtuse, so that the loading ishigher, and generally results in less controllableaction. Though hand tools are applicable to metalturning, and may in certain cases be used t oadvantage, they are rarely seen in modern machineshops, and in consequence few turners acquire theskill necessary to manipulate them properly. Ihave always believed that it is worth while t oobtain some practice in using hand tools, as job-bing work often involves some contours which aremore readily produced in this way than in anyother. One of my friends, whom I recently per-suaded t o try out ‘hand tools f o r turning machinetool handles, was agreeably surprised by the easewith which pleasing contours and high finish canbe obtained by their aid.

Special hand toolsSome turners employ a concave form hand tool

as shown in Fig. 3 for turning spherical or otherconvex contours. This certainly helps in produc-ing t h e shape, and avoiding the formation of hardlines or ridges, but the radius of the cutting edgeshould always b e greater than that of the sphericalcurve t o be produced. No attempt should bemade to use t h e tools as a forming t o o l in theaccepted sense; i t still calls f o r some manipulativeskill to produce the right ‘shape.

Left: Fig. 3.Hand turningtool withconcave edge.

A type of tool frequently recommended forspherical turning is that which consists of a tubewith the end ground at a bevel angle to form acutting edge. As steel tubing of a nature suitablefor hardening and tempering is not easy to obtain,this tool can b e made from silver steel rod, drilledconcentrically to a size which leaves a wall thick-ness of about 1/16 in., and long enough to give goodhand control, with or without t h e addition of anextension handle. The end is turned to a suitablebevel angle to provide a cutting edge, and afterhardening and tempering, this can be sharpenedby means of an oilstone while running in the lathe.

It is interesting to observe the effect of differentcutting angles, and their angular position in rela-t ion to the work, when using this t ype of tool.Sometimes the edge is bevelled internally at a

fairly acute angle (Fig. 4A), which is usually mostconvenient for working on soft materials, with thetool on an elevated rest, or inclined upwards sothat it is presented more or iess tangentially. Itthus works as an internally-ground gouge, whichis not the easiest of tools to manipulate on hardermaterial, even when the acuteness of the angle isconsiderably reduced. The tool is liable to becomeclogged with cuttings or swarf unless openings areprovided for its clearance.

Fig. 4: The use of tubular turning tools on soft andhard materials.

An externally-bevelled edge, at an angle to suitthe material to be turned, as in Fig. 4B, is gener-ally easier to manipulate, and may b e used oneither convex o r concave contours. There may bea case for a tool with both internal and externalbevel angles in certain conditions, but so far Ihave never had occasion to try it. Another usefulhand tool for spherical turning may b e made fromflat carbon steel by drilling a hole near the endand countersinking it to a suitable angle. Thisalso should be made long enough for comfortablehandling, preferably by providing a pointed tangto which a handle can be fitted. Tools of this typeare applied on the top of the work, with an ele-vated t o o l rest, out the cutting thrust is in t h etangential direction, so that for anything morethan a mere scrape, they call for more physicaleffort than tools in which the load is taken on thetool rest.

Gauging contoursWhen forming contours by “ freehand ” meth-

ods such as those described, form gauges or tem-plates may be used with advantage to checkaccuracy. Standard radius gauges will give a widerange of precise circular arcs, applicable to spheri-cal contours, but most of them only extend over90 deg., with straight tangential sides not applic-able to wider arcs. This prevents them being usedon work which is supported by a stem on the leftor rigbt-hand side, as it nearly always is. I havefound it worth while to make radius gauges asrequired by drilling and reaming holes in thinsheet metal and cutting away the unwanted pant;they may be adapted to check the shape of theneck or stem of the sphere, and also to cope with

426 MODEL ENGINEER 5 May 1967

TANG TO FIT ’HANDLEI

Fig. 5: Flat tool for ball turning.

more complex shapes. When using a template, asheet of white paper, or other reflector of dif-fused light, should b e placed behind the work, sothat the slightest deviation in its contour canreadily be detected.

I TEMPLATE )

Fig. 6: Use of template to check contour.

In industrial production, spherical and othercontours are often turned by means of form toolsmounted in the slide rest in which the cutting edgeis of negative shape to that of the article to beturned. The extent to which such tools can beused in jobbing work, however, is rather limited,primarily because accurate forming of the edge,by grinding or otherwise, is a difficult operation,and one tool will only produce one contour. Pre-cision of circular form is only possible at an exactradial setting, and allowance must be made forrake and clearance angles. If the tools are of sub-stantial width, the cutting load is very high, andwhen used in light la thes they are liable to chat-tering or digging in. For certain small componentswhich have to be produced to uniform shape andsize in quantities, form tools can be usefully em-ployed, and undoubtedly enable time to be savedcompared to other methods, but apart from thisthey have little application in the small workshop.Small form tools are generally made from flat toolsteel, which, after annealing, can be filed to therequired shape, and honed to a keen edge before

MODEL ENGINEER 5 May 1967

hardening and tempering. When resharpening isrequired, they should b e ground on the tap faceonly. For heavier and more continuous use, formtools of circular disc shape can be turned to therequired contour and notched to provide the cut-ting edge. These obviously have a much longerlife than flat tools, as the cutting edge can beground away and re-set many times before theyare worn out. Success in their use depends uponselection of suitable tool steel and its heat treat-ment, also on producing the highest possible finishof the contour face.

Some small ball turning operations can bespeedily carried out by the use of tools whichform approximately half the contour at one set-ting. A single tool, with approximately a quarterof the contour formed on each side as in Fig. 7,can be used, and as it is not (practicable to turnthe Ieft-hand side of the ball right to the centre,the right-hand side of the form tool is shaped toallow for leaving a certain diameter for finishing,or to provide a supporting stem if required. Thecircular arcs of the cutting edge can be machinedto shape by end milling cutters of suitable diam-eter. For accurate work, both the lateral andradial settings for the right and left hand contoursmust be critically adjusted. Such tools are suit-able for producing contours on work up to about1/2 in. diameter on small lathes; beyond this, thebreadth of the cutting edges ‘becomes too greatfor smooth and efficient operation.

Generating spherical curvesWhen it is necessary to machine spherical curves

to very close limits of accuracy, such as for usein a smoothly articulating ball-and-socket joint,the only practical method of doing so is by trav-ersing the cutting tool in a circular arc so that itgenerates the contour, instead of copying a fixedform. It is thus possible to produce sphericalcontours of any size by radial adjustment of thecutting tool, and maintain precision in all cases.The principle employed is simple, involving someform of pivoted tool carriage, capable of partialrotation, and generally including a slide for radialadjustment of the tool.

Right:Fig. 7.HalfContourformingtool. COMLETED

427

In the past , lathes have sometimes b e e nequipped with built-in rotating tool slides, thoughmore often, separate attachments have been pro-vided for spherical turning operations. The speci-alised lathes made for ornamental turning abouta century or more ago were notorious for theirarray of ingenieous and complicated attachments,some of which, I am sure, were never fully under-stood or mastered by many of their operators.While most of these lathes and their galaxy ofgadgets are now obsolete, some of the principlesof mechanical1 movement which they employedhave been adapted for use on the metal turninglathes used for special instrument or so-called“ precision ” work. These include spherical turn-ing appliances, which have been made in a varietyof forms, either by manufacturers or individuallathe users. It was, I believe, claimed for onewell-known " universal ” lathe that with its specialmovements, one could not only turn a sphere, butalso cut a thread on it (or even inside it !) thoughjust why anyone should want to d o so is not clearto me.

A ball clampSome years ago, I was faced with the problem

of producing a practical form of ball clamp,which involved the need for turning both externaland internal spherical contours. As this was aone-off job, which was not likely to recur often, ifat all, I decided to make the simplest possibleform of appliance I could think of which wouldfulfil the requirements for accurate work. Thisconsisted of nothing more than a hand lever,pivotally attached to a bolt fixed on the cross-slide of the lathe, and provided with a tool-posthaving some latitude of radial and slewing adjust-ment for the tool b i t Despite its primitive design,this device was found capable of carrying out therequired operations satisfactorily and has sincebeen found useful for innumerable other jobs.

Among important factors in the success of thisor indeed any other spherical appliance are therigitdity of the pivot, and the elimination of bothend and side play. It is clear that unwanted move-ment must necessarily affect the position of thetool point and thus produce inaccuracy o f form.The pivot must be Iocated exactly under thecentre of the work axis both laterally and cross-wise to produce a true spherical contour in theright place. If the pivot centre is displaced cross-wise the work is no longer spherical.

In terms of solid geometry, the result will beeither an oblate spheroid or an ellipsoid, accordingto whether the pivot centre is displaced towardsthe front or the back of the work centre.

For this reason, the provision of some meansof setting the pivot in true centre location is, tosay the least, very desirable, though surprisingly

428

enough, this has been omitted in some of theelaborate appliances I have seen. To cope withthis requirements, the pivot T-bolt, which isanchored in one of the T-slots in the ‘cross-slide,is made hollow, and a close-fitted gauge pin isprovided which serves a dual purpose. When set-ting up the appliance, a true running live centreis fitted to the lathe headstock, and the pivotlocated from it with the aid of the gauge pin.Lateral adjustment is of course ignored at thisstage, as it will depend on the length and positionof the work piece, but the cross-slide is adjustedso that the gauge point coincides with the lathecentre point as closely as possible, using a lens t oassist visual observation if necessary.

Setting the tool pointThe gauge pin can also be used for setting the

radial position of the tool point, by raising it suffi-ciently to allow the distance between the cylindri-call part of the pin and the tool to be measured,by calipers or other means. It is, of course, neces-sary to allow for the diameter of the pin whichshould be a standard or, at any rate, knowndimension) by adding half this amount for ex-ternal settings, and subtracting it for internal set-tings. The need for this allowance can, however,be avoided b y cutting away the top end of thegauge p in to exactly half its diameter.

In using this appliance, the lateral position ofthe pivot must be adjusted to a distance equal tohalf the diameter of the sphere from the end ofthe work piece, assuming that a full curve rightto the centre is required.

The cross position of the pivot should be notedon the slide index or, better still, located by a limitstop. It is permissible to run the slide out fordealing with oversize work, but it must be set tothe predetermined position for finishing, and atno time fed in beyond it. For most purposes,spherical rather than diametral accuracy is themore important, except when matching to internalcontours is necessary.

There are obviously many ways in which thissimple appliance could be improved, and thesewill be discussed in subsequent articles. Fromexperience in its use, however, I think that themost practical improvement would be to increasethe bearing surface of the pivot, and in particularto eliminate end play at the thrust faces. Thediameter of the pivot bush flange might well beincreased, together with that of the retainingwasher on the top of the lever. A frictionalspring washer under the nut has been tried, butwhile this resists the tendency of the lever to moveup o r down, it does not positively prevent suchmovement. Some degree of friction does, how-ever, assist in hand control in traversing the tool.

To be continued.


ELEMENTS OF SPHERICAL TURNING

Part II Continued from May 5

W HETHER the device employed for producingspherical contours is simple or elaborate, the veryfirst principle to be observed in its operation is thelocation of the centre around which the generatingtool is rotated, b y means of a hand lever or othermeans. This was referred to in the previous issue,but in view of its importance, further explanationof it is given in Fig. 9, which shows the effect ofdisplacing the pivotal point i n either direction. Ifcarried to the extent indicated, the error is readilyvisible to the eye, but for work in which highspherical precision is necessary, even the smallestdiscrepancy may spoil the finished results. Forthis reason, some positive means of locating thepivot centre on the cross-slide or bed of the latheis an advantige, and in some cases may be ab-solutely essential.

OBLATE SPHEROID E L L I P S O I D

Fig. 9. Effect of inaccurate locat ion of sphericalgenerating tool.

I n general work, some method of radial adjust-ment for the tool point, to deal with varying sizesof work, is equally important; it is also an advan-tage to provide the smoothest possible feed move-ment of the rotating tool fixture. But it may bementioned that in commercial production, spheri-cal curves of fixed dimensions, over a limitedangle of arc, are accurately produced by methodsand appliances which do not provide either ofthese facilities. Components such as ball-and-socket joints for motor car steering and suspen-sion systems, in which smooth articulation is abso-lutely necessary are often finished to close limitsof precision by means of a hollow grinding wheel,running at high speed on an axis at 45 deg. to thatof the work, as shown in Fig. 10.

The spherical arc generated in this way is usu-ally limited to about 90 deg., which is sufficient


for the particular purpose, provided that the restof the surface is undercut or relieved, as indicatedin the example of work dealt with. In order tomaintain the uniformity of the spherical diameterdespite the inevitable wear of the grinding wheel,it is dressed, when necessary, on the front flat faceonly. The same generating principle could beapplied to the use of a cutting tool such as ahollow milling cutter. Increasing the angle of thetool axis in relation to the work would enable thespherical arc to be extended within certain limits,but it would always leave either a cone or a flaton t h e front end of the work. It is just as import-ant, however, to observe the first principle, inlocating the axis of the grinding or cutting tool sothat it exactly intersects that of the work.

Fig. 10. Method of generating spherical contour withhollow grinding wheel.

It has already been mentioned that many in-genious spherical turning appliances have beenproduced for use on the lathe, either by manufac-turers or lathe users. Fig. 11 shows a typical ex-ample of such a device, ‘based on, but not copiedexactly from, an exhibit seen many years ago at aME Exhibition. It is designed to be Mountedeither on the bed or cross-slide of the lathe bymeans o f two bolts, and it incorporates a gearedrotary movement, in conjunction with a radialtool slide. As an example of machine tool com-ponent design, it deserves commendation, and themechanical details are wel l made and fitted.

The worm gear provides an adequate turntablefor the radial slide, with a broad under-surface to


7

form a thrust bearing, with adjustment for endplay from below, in the recessed bolster plate. Aremovable gauge pin, cut away to the centre tofacilitate radial measurement, is fitted to thecentre of the worm wheel. The radial slidingmember is fitted wi th a gib for adjustment, and ismoved b y a feed screw with a balanced handle,hating an indexed sleeve, though the markingson it are not visible in the photograph. A lanterntype of tool post, with rocker adjustment for thetool height, is anchored in a T-slot which enablesthe latitude of radial movement t o b e extended.Smooth and steady motion of the rotary traverseis facilitated by the worm gearing, and the wormshaft is fitted with a flexible drive, so that it canb e operated from any convenient position, or evencoupled to a self-acting drive if required,

Fig. 11. A worm-geared spherical turning appliance.

ofttn found that this does not give sufficient clear-ance. The only thing to be done then is to slew thetool round to the left in the tool post, but thoughthis enables the arc of surface to be increased, itreduces the advantages of fitting a slide whichgives direct radial tool adjustment. It will be clearthat the movement of the tool slide is no longertruly proportional to the radial adjustment at theactual tool point.

As I have often had to carry out spherical turn-ing, in the course of solving those awkward littlemachining problems which are passed on to mefrom time to time, I have devoted a good deal oftime to the design of appliances which give maxi-mum facility without serious limitations for workof this nature. The photographs, Figs. 12 and 13,show one of the most useful of these which I haveso far been able to (produce. In the drawing, Fig.14. a few minor modifications have been intro-duced in order to simplify construction withmaterials most likely to be readily available.

The utility of this appliance, however, is res-tricted in certain respects, wherein its refinementsmay, En fact, constitute a handicap rather than apractical advantage. In the first place, the verticalheight of the bolster, worm gear, and radial slideimpose a limit on the size of the work which canbe machined on a lathe of small centre heightwhen using the appliance. Though it is likely thatmost of the spherical turning required will be onparts of small diameter, there are occasions whenlarge work may b e encountered, calling for themaximum possible clear swing. The space occu-plied by the worm gearing, etc., can then ill bespared, and the appliance is at a disadvantage,compared with one having a direct lever feed.

Most spherical turning operations, other thanthose which are limited to a relatively small arc,involve the need for rotating the tool as far aspossible to the left, to carry the curve right downto the neck or stem. When the work is held in anormal chuck, or between centres, the radial slideand its handle may not be capable of swinging farenough round without fouling the chuck or otherrevolving parts. I t is possible to make the handleremovable, which helps to some extent, bu t I have

478

Fig. 12. A spherical turning appliance with obliquewormshaft.

The base of the appliance is a rectangular pieceof 1/4 in. steel plate, scraped flat and true on bothsides. For mounting it on the cross-slide of thelathe, two recessed socket head screws are pro-vided, which engage with 1/4 in. B.S.F. tapped holesin a strip suitably shaped to fit the T-slots of thecross-slide. This method, incidentally, can be

Below: Fig. 13. Another view of the appliance show-ing method of adjusting the radial slide.


LANTERNTOOL POST GAUGE PIN I

FlYW FEED/BLOCK

I CiAP 3 T.F!I. Z-START ACME THREAD LOCK/ NUT PlVOi BUSH

/ SOT. 20D.P SPUR GEAR SPHERICAL TURNING ATTACHMENT

VERTICAL PIVOT TYPE

Fig. 14.

FEED SCREW

recommended for securing all sorts of attach-ments to Tslotted cross or vertical-slides. Itfacilitates quick setting-up and removal, avoidsthe need for loose clamps and bolts, or projec-tions which may get in the w a y of operations.Moreover, they reduce the risk of overstrainingthe T-slots, by providing increased bearing sur-face over their full length. The strips can easilybe made from mild steel bar b y milling or shap-ing; they should slide easily in the T-slots, withclearance on the top surface, so that they lie belowthe slide level.

allows the gearing to be operated with the mini-mum interference from other slide movements

A hole is drilled and tapped 3/8 in. X 26 t.p.i. inthe centre of the baseplate, which is recessed onthe underside so that a thin locknut can be fittedto secure the pivot bush when it is screwed in toeliminate end play of the rotating parts. As in theappliance illustrated in Fig. 11, worm gearing isused for the rotating movement, but instead ofusing a proper worm wheel, it makes use of a spurgear with straight axial-cut teeth. The reason forthis is to allow the wormshaft to b e inclinedupwards at a sufficiently steep angle to give ampleclearance for the operating handle, over the cross-slide. Whether this shaft is located parallel to thelathe axis, or at any other convenient angle, it

The gear wheel actually used for the originalappliance was one which happened to be avail-able, in the traditional model engineers’ “ treasurechest ” (some unkind people refer to it as the“ jackdaw’s nest “). But in order to cater for thosewho may not be able to obtain a gear wheel of thesame size and pitch, the drawing has been modi-fied to show one which may more readily be avail-able. Change wheels of 20 d.p., and with varyingnumbers of teeth, are used on several popularlathes, and are obtainable as spares from makersor dealers. The 50-tooth wheel shown is a con-venient size, but the exact gear ratio is not import-ant. Both sides of the wheel should be flat andparallel on their essential bearing surfaces; re-cesses, if they already exist in one or both sides,are permissible, and the teeth on the lower sideshould be relieved.

Worm detailsTo mesh with a 20 d.p. gear wheel, a worm of

6 t.p.i. will serve for the purpose in view, thoughit is not perfectly accurate and in fact, calcula-tions are complicated by the angle at which it is

MODEL ENGINEER 19 May 1967 479

presented to the wheel. Perfectionists w h o insiston everything being exactly right may find, afterworking out details to the nth decimal point. thatthe exact pitch cannot be produced by any meansat their disposal. I n the solution of many engi-neering problems, approximate accuracy is per-missible, and gives quite satisfactory practicalresults, with certain reservations where the utmostefficiency is absolutely necessary. I mention thisbecause theoretical errors in such matters areoften a happy hunting ground for valiant Knightsof the Slide Rule; I am not deriding their worthyand often necessary efforts in working out diffi-cul t problems in workshop mathematics , butsimple methods will often produce the desiredresults.

The term 6 t.p.i. defines the distances betweenthe crests of adjacent teeth, or in other words the“ apparent pitch.” This is also the true pitch forsingle-start worms, but for a two-start worm, thetrue pitch, or “ lead,” is 3 t.p.i. To cut this pitchon a lathe having an 8 t.p.i. lead screw, the man-drel/lead screw ratio needs to be 8:3. This canbe obtained by using a compound t ra in withdrivers of 40 and 50 teeth, and driven gears of25 and 30 teeth, or any combinations which willgive total equivalent ratio. The worm blank maybe made in a readily machinable metal, such asbrass, as its duty is quite light, and after turningand concentric drilling, may wi th advantage bemounted on a true-running mandrel to run be-tween centres. The screw-cutting tool is ground toAcme form (29 degrees included angle) and thewidth at the tip should fit the root of the gearteeth; this is the maximum allowable, as it isoften more oonvenient to use a narrower tool andfinish t h e thread by taking side cuts.

The lathe catchplate should b e fitted with t w oequal-sized driving p i n s exactly opposite t o eachother, for cutting two-start threads. It is mostimportant that the blank should be q u i t e secureon its mandrel, and the carrier must also be firmlyfixed, as shifting of either during the entire opera-tion would ruin the work. If desired, a narrowstraight-sided tool may be used to gash out t h ethreads before using the forming tool. In either

case, the first “start” should be cut almost to fulldepth in one position before engaging the carrierwith the other driving pin, for cutting the second“ start.” Finishing the threads may be done inalternate driving positions, at the same tool setting.

It may be mentioned that the angle of thewormshaft is determined by the pitch angle of theworm, which in turn is influenced by its pitchdiameter. With an outside diameter of 5/8 in., thepitch angle works out at approximately 10 deg. asshown on the drawing, but this may be subject toslight variation. It may be checked by meshingthe worm hard against the teeth o f the wheel, withits bearing bracket in position. Any discrepancyin the angle can then be detected easily and, ifnecessary, corrected by filing the underside of thebracket. A gap will need to be cut to clear theworm when it is located at approximately the levelof the worm wheel centre, but this is not at allcritical. The bracket is located to give close butnot tight meshing, and it is secured by two screwsfrom the underside o f the baseplate. An exten-sion sleeve is fitted to the wormshaft to carry thehandle well a w a y from working parts, and eithera ball handle, or any other type preferred, can befitted. Incidentally, t h e spherical curves of thehandle shown were generated b y its own aid, whilea temporary handle w a s fitted.

A short radial slide is secured to the top of thegear wheel, it is intended for adjustment, ratherthan traversing in the normal sense. Its feed screwis fitted at the side and does not project outsidethe turning circle any more than necessary. Theslide is slotted so that a gauge pin can be fitted inthe centre of t h e pivot bush, and when this is notin use, a cover plate, held in place by a singleknurled screw, can b e fitted to prevent swarfclogging the slot. A lantern type tool post, withthe usual rocker pad for height adjustment, isfitted t o the moving component of the slide.

The radial movement provided is relativelysmall, but is sufficient for its designed purpose,which is to cope with spherical work up to 1-1/2 in.diameter on a 3-1/2 in. lathe. The tool can be rotatedthrough a full circle without fouling of the movingparts, though this is never necessary or practicable.

HOLCROFT VALVE GEAR

Page 274, column 1, line 6. Should read : Page275, column 2. First equation should read :Thus when Angle A = 90 d e g. N o t 20 deg. 1.620

O F = __ X .00894 = .0037Page 275, column 2, line 9. Should read : 3.939

and 337+ deg. Not 339 deg. So OF = 1.620 + .0037 = say 1.624 in.

480 MODEL ENGlNEER 19 May 1967


Part I I I Continued from May 19

THE ROTATIONAL movement of the cutting tool, forgenerating a true spherical curve, may be in anyplane, provided that it obays the first principle,that the centre line of the pivot must exactly inter-sect the axis of the work. Many appliances havebeen made having a horizontal pivot set at thelevel of the lathe centres, and a tool head equippedwith convenient means of rotation and radialadjustment. This arrangement has advantages forcertain applications, not the least being that itspivot location can be made positive, thus elimin-ating the risk of spherical error. It can also bedesigned in a form which is easily and quicklymounted, without the necessity for removing anyof the normal slide rest fittings. The equivalent ofheight adjustment for the point of the cutting toolcan be obtained by the cross-slide movement(which should be locked or clamped during thespherical turning operation) and the lathe mandrelmay be run in either direction, according to thetool setting.

In the course of operations to which I brieflyreferred in the previous issue, 1 was led to experi-ment with various designs of appliances with aview to producing a quick-set type of fixture,interfering as little as possible with the use ofstandard tooling arrangements. The horizontal-pivot appliance seemed to offer a possibility ofachieving this purpose, as it could be mounted onthe back end of the cross-slide, well out of theway of any bu t the largest work, and instantlyready when required for use. I already had a reartool with a detachable head, somewhat similar t othat designed by Duplex several years ago, andthis w a s used for t h e mounting of the sphericalturning appliance.

In the original arrangement of this device, therotating tool head was mounted in a spindle whichworked in a horizontal bearing block, designed tobe mounted on the pedestal of the toolpost fixture,interchangeable with the existing tool heads, andclamped down by the T-bolt in the slot of the cross-slide. A short radial tool slide was attachedto the front end of the spindle, and a worm wheelagain found in the scrap box fitted to the rearend, for operation by a worm with a vertical shaftand ball handle. The spindle bearing in the blockw a s bored and reamed in situ from the lathechuck while clamped on the toolpost at right-


angles to its normal position. Its height, therefore,was positively set, with no possibililty of introduc-ing spherical error, and it was unnecessary toprovide a gauge pin in the pivot centre.

The radial tool slide was, as in the previousappliance, of dovetail f o r m , with adjusting gib,but the feed screw could be centrally located, asit did not have to dodge a gauge pin Neither wasit necessary t o fit a rocker pad under the tool bitin the lantern toolpost, as the cross-slide providedthe means of adjustment. A bracket was attachedt o the back of the pivot block to provide a bear-ing for t h e worm shaft, which was extended toa convenient height for operation by the ballhandle. This device, though it started as a make-shift, and never really attained the status of afinished design, served its purpose, and enabledme t o produce some accurate work with the mini-mum expenditure of time and trouble.

Ball handlesSome months ago, at a meeting of the Sutton

Model Engineering Club, a question arose aboutthe best method of producing ball handles formachine tools. This led to a discussion on spheri-cal turning generally, and I demonstrated theappliances I had made, together with examples ofwork produced by their aid. Most of the clubmembers considered that it was not worth whileto make an “ elaborate gadget ” ‘to produce oneor two ball handles, and this is obviously true interms of workshop economics, as applied to allkinds o f tool making. But there are many modelengineers who will spare no pains to make theirworkshop equipment as complete and versatile aspossible, irrespective of whether they have anyurgent need t o carry out highly specialised opera-tions. Very rarely are efforts in this direct ionentirely wasted, provided that t h e gadgets theym a k e are practical, and not so elaborate as toabsorb considerable time, which might better bedevoted to more worthy projects , or what isknown as “ real model engineering.” This is amatter on which e a c h individual must form hiso w n judgment and make his own decision.

Mr Norman Cohen (who, incidentally, has nowemigrated t o Canada) was interested in theseappliances, and after trying out the horizontal-pivot type, decide to construct his own version of

MODEL ENGINEER 2 June 1967 529

it. This proved to be well suited to his require-ments and he has made complete detail drawingsof his design; which he has given me permissionto reproduce. Not only is he a much betterdraughtsman than I am, but his appliance is muchbetter made and finished than mine, so I havedecided to illustrate it in preference to the originaltype, from which it differs in many minor details.

The general arrangement, next page, shows thecomplete appliance, and the location and assemblyof the parts, to which the various members areallocated; to avoid confusion, therefore, thedetails are given only their part numbers and notfurther figure numbers. Part No. 1 is the pedestal,Which is essentially similar to that of the Duplexrear too1post, and if this already exists, can beutilised without the need for a specially madecomponent. Cast iron is the most suitable mat-erial, but it could be made by brazing two piecesaf mild steel bar together if preferred.

The underside of the base is first faced flat andtrue, and it is then reversed, either in the four-jawChuck or on the faceplate, with the square partset truly for turning the register spigot and drillingthe vertical hole for the main T-bolt. This bolt,and the nut which holds the assembly down, is notshown in the drawings or photographs, but itspurpose and location are obvious. Nothing morethan the single b o l t is really necessary for securemounting, but a hole for a short bolt is providedin the base flange f o r further security, but moreparticularly t o prevent the pedestal shifting if themain bolt is slackened for adjusting the head. In-stead of two separate bolts, studs screwed at eachend may be fitted to a T-strip to engage the slotof the cross-slide, as recommended for the pre-viously described appliance. A l l finishing opera-tions, such as scraping of the base or seating sur-faces, which may affect the height af ‘the pedestalshould be completed before proceeding with othercomponents.

The bearing blockDuralumin is specified for the bearing block,

part No. 2, but other materials such as cast ironor bronze are equally suitable. Steel can be usedif a bronze bush is inserted to form the bearing.The base surface o f the block is stepped to pro-vide a square edge which fits against the machinedside edge of the pedestal, to locate the blocksquarely ‘to the lathe axis when fitted. The seatingsurface on t h e underside must therefore bemachined by milling or shaping. An alternativemethod, which may ‘be more convenient, is to facethe underside of the block right ac ross , at thesame setting as the drilling and counterboring ofthe vertical hole, and attaching an aligning strip,or fitting a dowel in the top of the pedestal, toprovide (positive square location. The drlling,

530

facing and counterboring to a close fit on t h ebase spigot can then be carried out at one setting,with the block set up in the four-jaw chuck.

The block may now be mounted on the pedestaland secured by the T-bolt to the cross-slide, withits left-hand face set parallel with the lathe axisfor centre-drilling, drilling and reaming t h e boreof the bearing. This is the critical operation whichdetermines the spherical accuracy of the workproduced by the appliance, therefore care shouldbe taken to see that both the pedestal and theblock are seated truly on the cross-slide, withoutany chance of grk or swarf being trapped betweentheir surfaces; errors in height are difficult tocorrect afterwards. When drilling the block onthe cross-slide, with tools held in the lathe chuck,there is a tendency for it to slew round under theapplied end thrust, and this should be preventedby backing the work up by a drill pad held in thetailstock. Two or three stages of undersize drillingare recommended and if a reaming size drill is notavailable, a 31/64 in. drill may be used, followedby a D-bit, which can easily be made from silversteel rod. The end faces o f the bearing should bemaohined by mounting the block on a stub man-drel.

The rotating slideSuitable materials bar the ratating slide member,

part No. 3, are brass, bronze o r cast iron. It isfirst machined flat and parallel on both sides bymilling, shaping or facing methods; the edges andends, though of no functional significance, mayalso be machined for the sake of neatness. Thedovetail slot was produced by Mr Cohen with theaid of a vertical-spindle machine, but it could wellbe carried out by end milling in the lathe, withthe work mounted on a vertical-slide. As shown,no provision was made for adjusting play in theslide, bu t this calls for meticulous fitting, andthere is plenty of metal to allow of interposing agib strip, with adjusting screws. The slidewayshould in this case be made 1/16 in. wider on oneside (not symmetrically), and a strip of steel 1/16 in.by 1/4 in. fitted, as for the slide of the vertical-pivotappliance. For drilling and tapping the hole 7/16 in.fine thread, to attach the slide to its spindle, itmay b e mounted on the faceplate, so that positivesquareness is ensured.

A piece of mild steel 5/8 in. X 7/8 in. X 2 in. longmay be used f o r the radial tool slide, part No. 4,which is machined to dovetail section by the samecutter as that used for the slot of the former com-ponent. This is most conveniently done beforestepping the top surface, but the 3/8 in. hole may bedrilled to take a bolt for holding the part on t h evertical-slide. I t is set horizontally, with the underface outwards, and the top and bottom edgesmilled at one setting by adjusting the slide move-

MODEL ENGINEER 2 June 1967

--._,* ,.:


The spherical turning appliance made by NormanCohen, with examples 0fa~c.111 handles made with its

ment. This will ensure that the two edges areparallel with each other.

To mill the slot for anchoring the toolpost, aspecial T-slot cutter is normally required, but itis possible t o file this out accurately enough forpractical purposes if the centre channel is first cutto full depth by a 1/4 in. end mill. The work ofremoving metal by filing may b e reduced if two1/8in. holes 3/8 in. apart are first drilled horizontallythrough the block, to partially form the base o fthe slot. Filing is also good enough to shape thestep of the slide, if most o f the unwanted metal iscut away with a hacksaw; bu t maohining facili-ties, where they are readily available, should

always be exploited to full advantage on work ofth i s na tu re . A ho le i s d r i l l ed longitudinallythrough the slide to clear the feed screw, the nutof which is formed in a brass plug 3/8 in. dia., fittedto the hole already drilled, and cross-drilled andtapped to line up with the screw. This provides ameasure of flexibility in alignment, as the nut canswivel or slide in the hole to adjust its position.The sliding surfaces of the two m a t i n g parts maycall for final fitting by scraping, or the use of afine file, so that they work smoothly together, andthe slide gib, if fitted, is then adjusted to eliminateany perceptible play.

The toolpost, part No. 5, may be made f r o m1/2 in. dia. mild steel bar, 1-3/4 in. long, faced squarelyon the end, and grooved t o fit the T-slot in theslide. It is then chucked in the reverse position,tapered as the end, and drilled and tapped to takea 1/4 in. BSF screw. The hole may be drilled downas far as the lower end of the cross slot, to reducethe amount of metal which needs to be removedin forming it. End milling, with the work held ina small vice on the vertical-slide, is the quickestmethod of doing so, but as an alternative, threeundersize holes can be drilled and filed out tomerge with each other in an elongated slot.

Also in mild steel, the sleeve, part No. 6, issimply a piece of 5/8 in. dia. bright bar, drilled 1/2 in.dia. through the centre and faced at both ends.Its length is not critical, as the tool position canbe adjusted by the cross-slide. A socketed screw,either with or without a head, can be used toclamp the tool bit, but there is much to be saidfor the good old-fashioned square-headed screw.made of carbon steel, and hardened and temperedon t h e point only.

Part No. 7 is simply a mild steel disc, drilledand tapped 7/16 in. X 26 t.p.i. (or to match otherparts). I t is used to back up the rotating slide, andprovide the utmost rigidity when locked up astightly to it as possible and further preventedfrom movement by a grub screw in the side.

The Queenof the South ”returns forfurther service.See page 527.

532 MODEL ENGINEER 2 June 1967


Part IV Continued from June 2

THE SPECIFICATION of the worm and wheel forthis appliance (parts 8 and 9) are subject tomodification to suit available material. Thereis considerable latitude in the size of these parts,and the ratio of reduction need not be strictlyadhered to. Gearing obtained from breakdownof “ surplus ” apparatus can, therefore, be used ifit is anywhere near the stated dimensions, and astraight-cut spur gear can be suibstituted for theworm wheel, if the wormshaft is set at the appro-priate pitch angle as in the vertical-pivot appliancepreviously described. A ready-made worm andwheel of 40: 1 ratio can be obtained from Bondso’ Euston Road (Cat. No. 7/33), the dimensionsof which are larger than those stated on the draw-ings but within the permissible limits. This par-ticular item has the worm machined integral withthe shaft; It is possilble that a worm bored to fita separate shaft may also be available, but if not,it is only necessary to modify the wormshaftbracket (part No. 10) and i t s location on the bear-ing block (part No. 2).

Mr Cohen decided to make the worm gearinghimself, and has taken pains to work out thedetails for machining the essential parts. There isa good deal of unnecessary apprehension aboutthe problems of cutting gears, in general, andworm gearing in particular. It is perfectly truethat if gears are required to run at maximum effi-ciency such as for power transmission, wheremaximum ratio of output to Input Is essential,both the dedgn and production must be accuratewithin very close limits. But when the object ofthe gearing is simply to translate the speed anddirection of control movement, and the duty i svery light, approximate accuracy of tooth formand other factors is sufficient for all practical pur-poses.

Most worm wheels are “ throated,” that is,made with a concave periphery to increase thecontact area of the teeth of the wheel in mesh with

Right: Set-up of worm wheel blank for gashing teeth.Below: Hob for finishing worm wheel teeth.

10 T.P.I. 29’lNC. FORM


the worm; but this is not absolutely necessary,and there are many examples to be found ofw o r m wheels with flat crowns in dividing gear andother machine tool applications. In any case, truecontact of the worm and the teeth of the wheelcan only be maintained over a limited width.Little if anything can be gained by making thethroat of the wheel engage more than 90 deg. ofthe worm thread. All throated worms must bedesigned to suit the diameter of the worms withwhich they engage, the concave radius of the

LA1

c

‘HE

VIEW FROM HEADSTOCK END

TWEEN CENTRES

lzz&-J PLAN VIEW


Mr Cohen’s spherical turning appliance. Note tkeworm gear.

throat being equal to .that of the root of the wormthread, plus clearance allowance. If the teeth arecut by a hobbing operation, the throat can be leftslightly oversize, as the hob will produce the exactcontour required.

It is generally advisable to make the worm first,as it can be used as a gauge for meshing with thewheel. The specified pitch is 0.098, and as theworm has a single start, t h i s is also the “ lead,” atwhich the thread i s cut. The nearest approxima-tion to this w h i c h can be cut o n a normal screw-cutting lathe is 10 t.p.i., which is close enough forthis purpose. A narrow gashing tool may b e usedfor cutting the thread nearly t o its full depth, buta careful ly ground f o r m t o o l with sides a t anincluded angle of 29 deg., and a width of 0.031 in.at the t i p is required. A narrower tool may beused if after cutting to full depth, the sides of thethreads are shaved to produce the width by ad-justment of the top-slide. The operation will befacilitated if a material o f good machining qualityis used, and ample lubrication by cutting oil orsoluble emulsion is supplied.

Having turned the worm wheel blank t o ex-ternal dimensions, it should b e mounted on amandrel for cutting the teeth. There are variousdivergent opinions on n o w this can best b e donewith limited equipment; It is often said that wormgears can quite easily b e cut by using an ordinarys c r e w t a p as a hob running in the l a t h e andmounting the blank to turn freely on a verticalspindle at centre height. When fed into cuttingcontact with the tap, t he blank will rotate a t its

612

natural rate, and cut the teeth progressively to therequired depth-at least tha t i s t he theory, and itis sometimes known to work, on worm ‘wheels i nwhich neither the pitch accuracy nor t h e numberof teeth are of very great importance.

I do not recommend attempting t o cut the teethin this w o r m wheel by t h e above method for thesereasons: f irs t it is not generally possible to en-gage the hob to full depth and, therefore, it wil lstart cutting on a larger diameter than the de-signed pitch circle. As a result, rotation of t h eblank a t its natural rate will tend to produce alarger number of teeth than required, and notnecessarily an exact number, so that there is a riskof double-tracking or uneven indexing of theteeth. The use of an ordinary screw tap of Whit-worth or similar thread form will produce teeth ofexcessively wide pressure angle, though this maybe acceptable for s o m e light duties. But the widefluting o f a standard tap may cause their threadsto lose contact wi th those of the blank for partsof the revolution, giving still further cause formis-tracking or mutilation of the teeth.

In my experience, and in that of Mr Cohen,the best way of cutting the teeth of worm gears,in t h e absence of proper gear hobbing machinery,is first to gash the teeth to near finished depth byindividual indexing, and then finish them by freelyfloating the blank against a hob of the same dia-meter, tooth form and pitch as that of the wormto b e used. The hob m a y therefore, b e cut at thesame setting as the worm, preferably in tool steelwhich can b e hardened and tempered, though mildsteel hobs, case-hardened, have been successfullyused f o r finishing one or two worm wheels.

The teeth of the hob, up to about six in num-ber, may be milled o r filed, with the cutting facesas near radial as possible. They should be keptfairly narrow, but a little in excess of the threaddepth; i t is not necessary t o space them evenly.Backing-off is hardly possible but a little easing-off of the crests of the threads, by honing, is help-ful, and t h e cutting faces should b e honed keenwith a slip stone.

In order to set up the blank for gashing theteeth, some form of indexing device is required,but this can be of a simple or even primitivenature, s u c h as a spindle on one end of which theblank can be mounted, and a division plate o r a40-tooth change wheel on the other. A detent orlatch to lock t h i s in the required positions is neces-sary, and the bearing for the spindle may be inthe form o f a block mounted horizontally on avertical slide. A single point cutter, fitted t o acutter bar bebween centres, may be used for thegashing operation; its form may be approximatelythe same as that used for screwcutting the worm,and the depth of cut, applied by the vertical feed,about 1/16 in. The spindle needs to be set with its

MODEL ENGINEER I6 June 1967

axis at the pitch angle of t h e worm, by swivellingthe base of the vertical-slide to a n angle of 2 .5deg. t o the square position, checked with the aidof a protractor from the edge o f the cross-slide.

Some care will be necessary t o set the positionof the blank centrally, both laterally and cross-wise, in relation to the cutter. Exact measurementmay be difficult, but after setting by eye as closelyas possible a shallow witness cu t may be takenon one tooth, and the blank turned into a positionwhere this is readily visible fo r inspection. Whenset, the saddle and cross-slide should be locked,a n d e a c h t o o t h in turn gashed by feeding t h eblank vertically, preferably t o a positive stop ifthis can be arranged. The cutter bar is then re-placed by the hob, the indexing gear is releasedso that the blank can turn freely, and the spindleset square w i t h the lathe axis. W i t h the hob fullyengaged with the indexed teeth, the lathe is runat moderate speed, and vertical f e e d applied untilthe teeth are formed to their finished shape. Asthe pitch of the worm is very slightly greater thanthat specified, the teeth of the wheel should be onthe minus side in depth rather than otherwise. It

will readily b e understood that t h e circumferentialpitch of the tee th is proportional t o the diametermeasured o n the working pitch l i n e

It may be thought t h a t a good deal of space hasbeen expended o n a component which is hardlyrelevant t o the main subject. The basic principlesof spherical turning do not directly involve wormgearing, and it is only incidental to the construc-tion of appliances wh ich have been described. Butadvice is often sought on methods of cuttingworm gears for use in various kinds of smallmechanical appliances, and it is hoped that thisdescription will b e found generally helpful. Thereare other ways o f m a k i n g worm gears, some ofwhich may produce higher accuracy or efficiency,but their design may involve t h e use of formulae,which is not easy to apply in practice w i t h equip-ment normally available.

The shaft to which the worm gear is fitted (partNo. 11) may be made from 1/4 in. b.m.s. if it canbe chucked truly for turning down and threadingthe end; otherwise it is better to use largermaterial and machine it all over between centres.

To be continued

JEYNES’ CORNER Continued from page 610

It will be seen that Bourne was experimentingwith liquid fuel as early as 1834; he also usedsteam power for riveting in this year. Here area few of B o u r n e ’ s early pa’tents :

1834 : Marine engine governor.1836: Single eccentric reversing link-motion.

This, however, did not reverse the lead.1837: S t e a m starting and reversing gea r f o r

marine engines. Fitted in SS Don Juan 1837 .1838: Marine return tube boiler, having forced

draught and superheater.1838: Steam donkey pump for boiler feeding.1852 : Impulse valves.1852 : Balanced crank throws.Bourne also initiated the superheating or re-

heating of the exhaust of the high pressure cylin-der of a compound engine, t o replace the heat lostby radiation, and produce more power in the lowpressure cylinder.

Among the books written by Bourne are: “ TheCatechism of the Steam Engine,” “ Recent Im-provements in the Steam Engine,” “ A Treatise onthe Steam Engine,” “A Handbook on ‘the SteamEngine ” and“A Treatise on the Screw Propel-lor.”

Bourne also had something to say about theterm Nominal1 Horse Power: “ The uncertaintyand varying character of the ratio subsisting b etween t h e actual and nominal power is a sourceof much perplexity, and proposals have beenmade in consequence to substitute some other unitfor the horsepower. In the average class of

MODEL ENGINEER 16 June 1967

A Fowler double-drum ploughing engine built i n1867. The late C. E. Shackle built a model of this

engine which is now in the Science Museum.

modern engines (1869) the actual horsepower maybe taken at 4 to 4-1/2 times ‘the nominal. The actualand nominal powers of engines, at first identicalor nearly so, soon began t o diverge; and in timeas the pressure of the steam was increased theactual power became twice greater than it hadbeen at first, while the nominal power being anexpression of the dimensions of the engine re-mained the same. The divergence however didnot stop here, but has gone on increasing, until inrecent engines, the actual power exerted has beenin same cases nine times greater ‘than is repre-sented by the nominal power.”

In conclusion, I would say that it at once be-comes apparent that confusion has existed on thissubject for over 100 years.

613

Part V Continued from June 16

THE thread should preferably be screw cut to fitthe rotating and lock ring slide neatly and accur-ately. It should be a close working fit in its bear-ing and adjusted to eliminate end play; if theworm wheel is keyed to the shaft as recommendedthe large headed screw (part No. 16) will provideendwise adjustment, and a fibre washer may beinterposed between the lock ring and the face ofthe bearing to enable slight frictional resistanceto movement to be obtained.

Modification of the wormshaft (part No. 13)and its bearing bracket (part No. 10) may becalled for if a different worm and wheel than thatspecified are employed, but the alterations areobvious, and the bracket only needs to be locatedagainst the back of the bearing block so that theworm is in close mesh with the wheel eliminatingbacklash as far as possible consistent with smoothand easy working. In the original version of thisdesign, I used a single bearing with a long sleeveextending upwards, instead of a gapped bracketas shown, the .object being to use a worm whichwas machined integral with its shaft. If the speci-fied arrangement with the worm a light press fiton the shaft and grub screwed or cross pinned toit, is employed, the gap in the bracket should bemade to fit the length of the worm closely, with-out end play.

The ball handle (part NO. 14) was machinedwith the aid of the appliance together with othersimilar pieces of various sizes, such as those seenin the photograph in the previous issue. Somefurther information on the {methods employed forturning these parts will be given later. Crossdrilling and tapping the handle to fit the worm-shaft and the crank may present practical prob-lems and, in the first case, it will be found worthwhile to set the piece up on the four-jaw chuck sothat the middle ball runs truly. Before starting thehole wi th a centre-drill, a light facing cut shouldbe taken t o serve as a “witness,” when any errorin setting up will immediately be apparent, andmay be corrected by adjustment of the chuckjaws. The cross hole, in this particular instance,should stop short of passing right through the ball,for the sake o f neatness, and a plug tap is thenused to produce a blind thread to screw tightlyon to the wormshaft. A short piece of 1/4 in. rod,similarly threaded on the end may be screwedinto the ball to assist in the orientation of thepiece when drilling and tapping the hole in the end

MODEL ENGINEER 7 July 1967



ball from the other side for fitting the crank. Ifthis is not done, it may be found difficult to ensurethat the two holes are in the same axial plane, andthe finished result may be rather unsightly, to saythe least.

The rotating tool slide is moved by means of asmall feed screw (part No. 10) which is quiteeasily machined from the solid, but it is permis-sible to make the head separately and pin o rotherwise secure it to the shank. Instead of screw-ing it 2 BA, a 3/16 in. X 40 t:p:i. thread may beused which, besides being more convenient forscrewcutting, would enable the head to be indexedin 2.5 divisions, each giving 0.001 in. incrementsof feed. But, as I have pointed out, this only givesaccurate measurement of the work being pro-duced when the tool is applied in a truly radialdirection, which Is not always possible, because oflimited clearance between the moving parts of theappliance and the lathe chuck when in action.

A groove is turned In the head of the feed screwto fit closely over the thickness of the keep plate(part No. 1.5) which is attached to the end of therotating slide (part No. 3). This has a U-shapedgap to fit the groove diameter of the screw, thusproviding positive location in both directions,without the need for a separate thrust collar.After drilling and tapping the fixing screw holes,the plate is located on the end of the rotating slide, with the feed screw In posibion, in the tapped holeof the radial slide (part No. 4) and the tappingholes spotted through, prior to drilling and tap-ping.

Ball and socket jointsAlthough the need for precise spherical accuracy

is most important In parts such as ball and socketjoints, which need to articulate over a universalangular range, many occasions occur in generalworkshop practice where it is worth while to setup a spherical turning appliance to ensure neatnessand genera1 accuracy of the finished component.These include several kinds of fittings for machinetools and attachments within the scope of tool-makers and “gadgeteers.” The two examplesshown are typical and do not call fo r explanation.Balanced handles are almost universally employedfor slide rest and machine table feed screws nowa-days, having superseded the plain crank handlesWhich were once common. Their primary object,when first introduced, was to avoid risk of in-

641

n - I

advertent movement under the effect of gravityand viabration, and this is still of some importancewhen the handles are large and heavy. Exactbalance is rarely necessary for small handles, butthe shape has become very popular because ofease of manipulation and pleasing appearance.

Clamping levers for turrets and other fixturesare also commonly made with ball ends, to har-monise with the handles, besides combining maxi-m u m strength with production fac i l i t i e s . Theshape and proportions of the examples shown arebased on examination of a number of parts seenon actua1 machines. Although standardisation ofthese parts by firms specialising in their produc-

BALANCED HANDLE FOR SLIDE REST

CLAMPING LEVER FOR TURRET

tion, h a s been introduced, there is still a good dealof variat ion in detai ls . The balanced handleshown has the large ball twice the diameter of thatof the small end; the middle ball is intermediatein size, though in some cases the method of attach-ment t o the feed screw, o r t h e f i t t i n g of an indexdial may cal l for modif icat ion of i ts s ize , andsometimes its shape as well. Smal1 handles areoften screwed on to the end of the feed screw andlocked by means of a back nut, In which case thecentre hole, as shown, does not pass completelythrough the ball but in other cases the handlemay be secured by a nut or retaining screw onthe outer end.

General proportionsThe hand crank has a length approximately

equal to the radius of the handle (i.e. half thelength [between ball centres) and a diameter some-whait less than that of the small end ball It isnecessarily made as a separate part and pressedor screwed i n tightly. Unless a contour-formingor copying appliance is available, its shape is bestproduced by means of hand tools. It is worth

while to take pains in shaping it properly, not onlyon the grounds of appearance, but also for com-fort in handling. One often sees rough and slov-enly shaped handles, neitiher the look nor the feelof whlich give joy to the operator. The shank ofthe balanced handle should have a taper of about2 to 23 deg. , to ta l length between ball centresabout four times that of the centres ball, andmean diameter about half that of the same ball.

For a clamping lever 2-1/4-in. long between ballcentres, the large ball may be made 11/16-in. dia., andthe small ball 7/16-in. dia.; other sizes in propor-tion. The taper may be 2 to 2-1/2 deg. as before,with a mean diameter of about 9/32 in. In mostcases levers of this type are cross drillled andtapped at an angle Which is convenient to giveclearance over the fixture and fittings to which itis applied, and unless the length of thread on theclamping bolt necessitates drilling right through, ablind hole is preferable for the sake of appear-ance. The angle of the cross hole in this exampleis 15 deg., and for drilling the hole, the large endball can be held in the three-jaw chuck, protectedby a sllip of thin soft metal, and inclined back-wards at the required angle; this is not at allcritical unless demanded by specification, or toensure uniforrnity in a number of similar com-ponents.

It may be observed in passing that solid integralclamping levers of this kind are not only neat, butalso provide maximum strength and rigidity ,fortheir intended duty. Levers made in two pieces,with the handle screwed or otherwise attached tothe centre hub or nut, obviously have their torqueresistance limited by the strength at the point ofattachment, which is often relatively weak. Avariation of the lever shown, which may be pre-ferred if the radial length is short, is to shape theshank like that of the hand crank of the (balancedhandlle, instead of with a ball end. The area ofthe clamping surface at the face of the cross holemay be considered too small for ample bearing,but a loose thrust washer may be fitted, and forsome purposes it may be advantageous to makethis with a spherical under surface to assist self-alignment of the fixture.

Machining procedureIn these, and most other turning operations of

a similar nature, roughing out to within about1/32-in. of finished shape and size is generally advis-able. The slide rest tools may be manipulated forpart forming of the contour, but, as the sphericalappliance is capable of taking reasonable cuts, itis often sufficient to take the corners off wiith a45 deg. chamfering tool. For the balanced handlethe taper shank can be turned to finished size, orleaving only a very small amount for subsequentfinishing; one advantage of the horizontal-pivot

642 MODEL ENGINEER 7 July 1967

-

_--___-- -,----OFFSET BAR FOR CONVEX CONTOURSj,‘. .y*

.-_ ‘fI \ -2.4

SADDLE FREE TO SLIDE-

D ONSLIDE RADIAL BAR

FECTIVE RADIUS

GENERATING LARGE-RADIUS CONTOURS.

appliance, when fitted at the back of the cross-slide, is that the normal front too1 can be retainedin position without disturbing the angular setting.A tool having angular sides at about 60 deg. inclu-sive and a small radius on the nose is advised forturning the shank, as sharp internal corners makeblending-in of contours more difficult. The twoparts of the shank should be machined at the sameindex setting, so that they form in effect a con-tinuous taper, interrupted by the middle ba1l.

While it is not impossible to form most of thecontours on work mounted between centres, it isgenerally easier to mount the work in the chuck,without back centre support, and machine asmuch of it as possible ibefore parting off. Thesmall end ball should be formed on the outer endand, if the amounlt of overhang should be con-sidered excessive, a little extra length should beallowed so that a small centre may be drilled totake the back centre; this length is turned downto a small diameter and subsequemly parted off.Sometimes a small axial hole in the end ball ispermissisble, such as to take a grub screw to lockthe hand crank. If the work can be chucked truly,the turning may be done piecemeal, with onlysufficient projecting for the particular part of theoperation in hand. After parting off, the large ballmay be finished by chucking over the middle ballagain protecting its surface with a soft metal slipor ring and pushing it back as far as it will go inthe chuck. It is generally possible to get within30 deg. of the centre in the spherical forming ofthe large ball, so that not much is left to bemachined in the finishing operation. Mention has


been made of the cross drilling operations on thesecomponents, which involve some problems, as itis not easy to drill exactly through the centre ofa ball, or to ensure that the hole is drilled at thecorrect angle, but detalled instructions on thesematters are somewhat outside the scope of thepresent articles.

Large radiiLimitations in the radius of adjustment for the

tool in any of the spherical turnling appliances des-cribed may make them unsuitable for dealing withcontours of specially large radius. It is, howeverpossible to cope with these by means of simpledevices which follow the same basic principles.Optical instrument work for instance, often in-volves the need to machine concave mirrors, alsolaps and moulds for lens grinding, which have alarge radius of curvature. The method is shownin my drawing, in which a radius bar pivoted at itstwo ends to bolts fixed to the cross-slide and thelathe bed respectivly, is employed. In order toproduce a true spherical curve, the latter pointmust be exactly under the lathe axis, and the toolpoint must be set to coincide with lathes centre-linewhen in the middle of the work as shown.

With the saddle free to slide and the cross-slidetraversed by its feed screw in the normal way, theswing of the slide bar will cause the tool to movein an arc and produce a contour equal in radius tothe length of the bar between its eye centres. Ina variation of this device, the cutting tool may becarried in a fixture attached directly to a dieblock fitted to a slot in the cross-slide. The effec-

643

tive radius In this case will then be equal to thedistance from the pivot on the lathe bed to theactual tool point, and therefore will be more sus-ceptible to variation or adjustment than whenusing an articulated bar of fixed radius. Convexcurves may be more difficult to produce in thisway as it may be impossible to pivot the radialbar on the lathe bed at the required point unlessthe work is overhung a long distance from theheadstock. In this case, it is possible to use anoffset radial bar at the back or front of the head-stock, provided that this is set exactly parallel tothe lathe axis when the tool is in the centre of thework. Ihis arrangement is indicated by dottedlines.

Curved profilesSometimes spherical turning appliances can be

used with advantage to produce-paradoxicallynon-spherical contours involving circular arcs.This cannot be done with the horizontal pivotappliance having a fixed centre height, but thesimple vertically pivoted lever device, which hasthe widest range of adjustment for either internalor external curves, is easily adapted to this kind ofwork. In my next drawing, the convex curve onthe end of the work is produced by setting the tool to a large radius from the pivot A, and the con-cave curve by setting it to a large radius frompivot B.

PIVOT A-Yk

G E N E R A T I N G N O N - S P H E R I C A L C O N T O U R S

This principle could be applied to forming thehand crank for the balanced handle, but manyturners would conslider it unnecessarily compli-cated. it was, however, employed to advantage inproducing a master pattern for use on a copyinglathe, where both the dimensions and curvaureswere specified within close limits. Where the con-vex radius is too large to be obtained with anappliance mounted on the cross-slide, the toolmay be controllled by a radial bar having itsstationary pivot mounted at the back of the lathebed, and the other on the cross-slide. In this case,the cross-slide must be set to move freely, and the

saddle traversed in the nonmal way, using self-actif desired. This arrangement was employed in amunition factory in quantity production of largeshells for forming the nose contour to fine limits.

A compound turning applianceWhile on the subject of appliances for spherical

turning, mention may be made of an ingeniousand elegant device built by Mr. John Pickles, asa part of a set of equipment for ornamental turn-ing. It is intended to be mounted directly on thelathe bed, and its movements provide for manyother operaltiion’s in addition to spherical turning.The square base carries a turntable with a “ tan-gent wheel ”or worm gear, the movement ofwhich is controlled by a crank on the shaft extend-ing to the lef t . With the centre of the basemounted exactly under the lathe axis, sphericalcurves can be produced by a cutting tool mountedon the radial slide, the handle of which extendsto the right. Superimposed on this slide is anothercomponent capaible of both rotating and radialsliding movement, and also incorporating a bear-ing for a horizontal spindle. This is fitted with agrooved pulley on its front end, so that it can bedriven from an overhead shaft, and a cross-slidecarrying a socketed tool holder. it is thus possibleto use the spindale for drilling, grooving, flutingand eccentric cutting or “ engine turning,” on flat,angulacr or spherical surfaces of work held in thelathe chuck.This appliance, together with a geo-metric chuck, was awarded the Bowyer-Lowe Cupat the 1958 M.E. Exhibiition.

A spherical turning slide rest built by John Pickles.

Allthough complex movements such as providedfor by this appliance are rarely required in mach-ining colmponents in general engineering, theydemonstrate how inexhaustible is the range of

Continued on page 652

644 MODEL ENGINEER 7 July 1967

SIMPLEXContinued from page 649

rather more than 1-1/4 in. wide, machine this, usingthe four-jaw, or by milling to exactly 1-1/4 in. wide,or If the frame slots have already been cut andfinished-to a tight fit in the horn slots. Drill allthe holes in the horns, noting that the top andbottom holes must not be too close to the ends orthere will1 be no room for the bolts holding thetop and bottom stays. Jam the “gauge” into thefirst horn slot and clamp the horns, in pairs, toboth the frame and the gauge, using two small,and one medium-sized clamps. Run the drillthrough the holes in fhe horns and drill theframes. Remove aI1 parts and clamps, take off anyburrs, and countersink the outside of the holes inthe frames. The rivets required here are 1/8 in. X1 in. iron snaphead, and these should be just aboutthe right lerrgth as bouglht.

Feed pumpThe most simple method possible to get water

into the boiller is probably the single-ram eccentric-driven pump, or an externally fitted crossheadpump. I would have liked to have shown a cross-head pump, similar to that on Rob Roy but theWalshaerts valve gear rather gets in the way. Itwould just be possible to sling a crosshead pumpof small bore underneath one of the cylinders, anddrive it by an extenslion of the drop link or by aseparate link bolted to the crosshead. The snag ofthis arrangement is the large offset necessary, in-troducing an unpleasant couple. This wouldquickly cause wear on the crosshead slippers andthen on the piston rod gland.

If we use a single ram eccentric-driven pump

SPHERICAL TURNINGContinued from page 644

operations which can be carried out on the lathebeyond normal turning as generally understood.Lathes designed specially for ornamental turninga century or more ago were often equipped witha galaxy of appliances, few of which, I imagine,were ever properly mastered by a majority ofthose who used the lathes.

There are many possible variants of the spheri-call turning appliances I have described, all cap-able of accurate work provided that they followthe basic principles. I have seen a simple hori-zontal-pivot attachment used on a capstan 1athefor production work, the spindle worked in asquare block held in the tool post of the cut-offslide, and the cutting tool, adjustably mounted on

653

between the frames, the large bore and strokenecessary to ensure that the pump is alwaysmaster of the boiler becomes rather a nuisance,and might cause jerky running at low speeds. Butreading through a recent article on pumps byEdgar Westbury gave me the idea of using adifferential double-acting pump, so that only oneeccentnic and eccentr ic rod and s t rap will beneeded.

To be frank, I am not too certain of the effec-tive output of the pump shown, so I have pur-posely arranged for a fairly large bore and stroke.Experienced builders who intend to fit an injectorin addition may be well advised to make thethrow of the eccentric 7/32 in., ratiher than 1/4 in.,which will reduce the anguliarity of the eccentricrod quite considertibly. 1 should however explainthat as the pump has not .been drawn exactly toscale, allthough the dimensions are correct, theeccentric rod appears shorter in proportion to theother components than it really is.

The pump is bolted to a cross-stretcher cut down3/8-in. thick b.m.s. by four 2 BA hexagon-headedbolts, though Allen screws would be preferalble asthere must be no chance of the pump workingloose in service.

This crossstretcher is in turn bolted to theframe.s by three 4 BA screws each side. That onthe right in my drawing should be countersunk, toclear the driving wheel, but the other two may beturned bolts, to ensure a good fit in the frames.In addition, after the bolts have been tightened up,a 1/8-in. silver steel dowel pin should be put in, agood fit In frame and stretcher, to take some ofthe thrust of the ram off the screws. There is justroom for this between the two bolts.

To be continued

its from flange, was presented tangentially to thework. Rotation of the tool head was arranged sothat it could be linked to the capstan for self-acting feed.

Modern users of lathes are generally satisfiedwith straight forward slide movements, which willgenerally cope with most machining requirements.Many readers may say “these gadgets do not con-cern me, I shall never need to do spherical turn-ing or other fancy work!” But one never knowswhen the need for special operations may arise,and when it does, it is allways urgent. The “ com-pleat ” turner (to borrow a term from Izaak Wal-ton) should always strive to attain a comprehen-sive understanding of both the possiblities and thelimitations of his machine. It was said by JoshuaRose, a well known 19th century writer on work-shop subjects, that “ he who is master of the latheis master of the entire mechanical world.”


Documents

ELEMENTS OF SPHERICAL TURNING - Model Engineer