How to make  spur and bevel gears  using milling machine and shaper,  the set-up for pinions, tapered packing plate and blank holder.
 

(This article is from Model Engineer Magizine, I reccomend it to all
hobby machinists)


Having built four locomotives -Rob Roy and Jubilee in 3 1/2 in. gauge

and the two Kettering Furnaces No. 3 in 7 1/4 in. gauge - I felt that a change

would be as good as a rest and looked for a different type of project. Initially a clock

was favoured and several designs examined and discarded mainly because of the

woodworking involved. I am one of those persons who picks up a nail to hammer

into a piece of wood only to find that the timber has split.

Thoughts then turned to a traction engine, machines which have held a fascination

for me since early childhood; indeed I had once gone as far as building the boiler

for an Allchin 'Royal Chester'. However, I had disposed of the castings some time ago

as I felt that a lot of the work was too fiddly for my deteriorating eyesight.

After examining a number of general arrangements the Fowler R3 two cylinder

compound, three speed engine offered by Messrs Plastow was decided upon. Drawings

for this engine are offered in both of its incarnations, as a Road Locomotive and

Showman's Engine. Whilst being a necessarily complex design this was of no matter

as I was looking for an 8-10 year project. This is the drawing which was published as Fig. 1 of the original article by Base Circle, see text for an explanation. Coincidentally Mark, the young man who had assisted me with the first steaming of the No. 3's asked to participate andbuild the Road Locomotive version as I had decided to tackle the Showman's

Engine. This is proving to be a most satisfactory arrangement as Mark has

some very useful skills and has undertaken to obtain materials, it will be very useful to

have an extra pair of hands at boiler making time.

Examination of drawings obtained from Messrs. Plastow indicated that a number of

gears would need to be produced in a range from 13T to 96T in 10 DP plus bevel gears

and pinions and thought was given as to how these could be made. One of my most

used workshop tools is an almost complete set of 'Model Engineer's' and recourse was

had through the use of my self compiled master index to see how past contributors

had tackled the problem. In Model Engineer, 14 September 1950 an article by Base Circle describes a simple and elegant method of coupling a gear using a simple tool to generate teeth. As Base Circle points out this produces teeth of correct involute form whatever the

number being cut. Being the possessor of a small 'Perfecto' motorised shaper this

appeared to be the way to go, especially as my old WWI. Cater and Hakes No. 2

horizontal milling machine has insufficient capacity to handle the larger gears using

involute gear cutters. Fig. 1: Machining the blank for the 96 tooth gear on the milting machine. Note the use of adaptors for the holder.

MODEL ENGINEER 20 JANUARY 1989

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Part II (conclusion), from page 112 Part I of this mini-series told how the Author

adapted a method which was described in Model Engineer, 14 September 1950, to

produce gears for his Fowler traction engine. He described how he got over problems of

rigidity which arose. This time he goes on to discuss the methods adopted to produce the

spur and bevel gears for the model. He then talks about making keyways for various components.

ith the spur gears successfully completed attention was

focussed upon the bevel gears and pinions for the compensating gear

(differential). Messrs Plastow's design calls for two 44T gears and two 9T pinions per

engine. He offers a specialised gear cutting service. However, having satisfied one minor ambition by producing the spur

gears, I wished to pursue the project to completion by producing these more complex

items. Back issues of the Model Engineer had provided a number of solutions, the most

attractive being a method of cutting to a parallel depth. The method has advantages

over conventional methods in that neither specialised machinery nor non-standard

cutters are required. Mr. Minchin, who described the system in Model Engineer, 15

November 1964, claims that power transmission is 80% of conventional gears, but for my purposes (and I suspect most others) this small loss was of no moment. Again, this seemed to be a simple and elegant solution requiring only the use of a

horizontal mill. Machinery Used As mentioned in the previous article I

own an old Carter and Hakes No. 2 milling machine. This would seem originally to

have been belt driven from an overhead line shaft and is clearly a production

machine. Judging by certain wear it has had a hard life. It came into my possession

in the late 1960's for 5 and has been given a new mandrel, extended longitudinal

table traverse, chain drive and a graduated Australian reader C. Bamford decided to build traction engines. He chose to make all the spur and bevel gears first, using milling machine and shaper Above: Fig. 8 shows the set-up for pinions, tapered packing plate and blank holder. screw for vertical travel. An oddity is that both longitudinal table travel and vertical

feed can be traversed by a lever operated rack and pinion. Four speeds are obtainable, through what I understand, is a 1929 B.S.A. motor cycle gearbox. Reading and re-reading the article I decided upon a method somewhat different than that described by Mr. Minchin in that he used a dividing head to index the blanks, using vertical feed for the gears.

My milling machine has no power feed for vertical travel and so I was looking for a

method to utilise the table feed. Also, although owning a Myford dividing head, I

felt it might not be rigid enough for the very heavy work envisaged.

Having decided upon a method, I was fortunate in being able to contact Mr.

Minchin and he most generously amplified his article and provided drawings of blanks

and cutters. The cutters proved to be of quite simple shapes and he suggested silver steel, which in the event proved very satisfactory.

Fig. 8 shows the basic set-up. A packing plate machined to a taper of 11 1/2 deg. to

match the pinion and bevel angles is bolted square across the miller table. For the pinions the fixture previously used to rotate the spur gear blanks was fastened to the packing plate and the blank bolted thereon. Indexing Arrangement Fig. 9 shows the indexing arrangement.

An odd length of angle is bolted to the rear of the blank holder and a circle of plastic to

the shaft. Nine indexed holes are drilled The method calls for a central roughing cut and one each side with the cutter offset by 25% of the pitch with the blank rotated in the same direction also by 25% of the pitch. This was achieved by indexing the plastic direct to holes above and below the gashing index hole drilled through the angle. Spacing was taken from indexed holes drilled when dividing was carried out. Movement of the cutter was effected by

means of appropriate washers upon the mandrel. Cutting proved to be quite free in the

 

 

 

MODEL ENGINEER 17 FEBRUARY 1989

 

 

reversed for this blank and the table clamp adjusted to withstand climb milling forces.

Again, cutting was quick and clean. One of the most serious problems was to

locate the blank centrally beneath the cutter and this was achieved using square

and caliper. The appearance of teeth being offset in the figure is an optical illusion. Fig. 11 shows a set of completed gears, which were a pleasure to make, and which mesh perfectly, although by the time both sets (for two engines, remember) had been Above: Fig. 9 shows offset cutting. The cutter holder is offset by removing the washer, pinion rotated completed a certain tedium had set in. This and indexed to the top hole in the angle. Below: Fig. 10 shows climb milling the bossed bevel gear. The indexing details are apparent in this was dispelled by passing to the keywaying view, as is the simple tool shape of 20 deg. included. operations. Above right: Fig. 11, one set of completed bevel gears and pinions. cast iron used, at the slowest table teed. stub of metal to locate the gear bore with Keyway Cutting

Fig. 9 also shows the first offset cut. Both an extended stud, nut and washer for A number of the gears slide upon four washers are to the left of the tool holder clamping purposes. A countersunk screw splines and others are keyed to their shafts and the pin is indexing in the upper hole in and nut affixes the blank to an indexed disc by the usual single key, and a method was the bar. Also visible is the simple tool and holes, at suitable spacing, are drilled devised to use the shaper to form keyways holder and the indexing holes. into the packer. Fig. 10 shows the method For cutting the bevel gears the angled and the simple tool used. Because of the in the gears. The angle plate previously described was packing plate was modified to accept a boss in one gear, the table feed had to be set, short angle vertical, upon the Myford boring table and a 1.25 in. dia. hole bored through. A turned plug was indexed for four 1/8 in. dia. holes, clamped to the bored angle plate face and the holes transferred

to the plate. Gears requiring four keyways had a single hole drilled at the correct pcd and a

stub of 1/8 in. dia. metal pressed in lightly, locating in each of the four holes in turn

thus locating keyways accurately. Each gear to be keyed was arranged to have a 1.248 in. dia. spigot for locating into the bored hole, the spigot being removed as a final operation if not required. Bore diameters varied from 0.6 in. to 1 in. and key widths from 5/32 in. to 3/16 in. A suitable tool holder was affixed to the face of the vertical slide of the shaper which has

a convenient 5/8 in. diameter hole being utilised for the purpose.

Once the cutter was centralised, the table was locked and cutting proved quick and

easy, depthing was by using the slide's graduated collar. Unfortunately the

photograph taken of the operation met with a mishap, but the concept is simple

and, I am sure, well known to many. In conclusion, a most rewarding project

which gave me much pleasure and which taught me a lot, not least that the methods

most generously described by Base Circle and Mr. Minchin work, and work

well. My thanks to them both. MODEL ENGINEER 17 FEBRUARY 1989

 

 

 

 

 

 

 

 

 

Having built four locomotives -Rob Roy and Jubilee in 3 1/2 in. gauge

and the two Kettering Furnaces No. 3 in 7 1/4 in. gauge - I felt that a change

would be as good as a rest and looked for a different type of project. Initially a clock

was favoured and several designs examined and discarded mainly because of the

woodworking involved. I am one of those persons who picks up a nail to hammer

into a piece of wood only to find that the timber has split.

Thoughts then turned to a traction engine, machines which have held a fascination

for me since early childhood; indeed I had once gone as far as building the boiler

for an Allchin 'Royal Chester'. However, I had disposed of the castings some time ago

as I felt that a lot of the work was too fiddly for my deteriorating eyesight.

After examining a number of general arrangements the Fowler R3 two cylinder

compound, three speed engine offered by Messrs Plastow was decided upon. Drawings

for this engine are offered in both of its incarnations, as a Road Locomotive and

Showman's Engine. Whilst being a necessarily complex design this was of no matter

as I was looking for an 8-10 year project. This is the drawing which was published as Fig. 1 of the original article by Base Circle, see text for an explanation. Coincidentally Mark, the young man who had assisted me with the first steaming of the No. 3's asked to participate and build the Road Locomotive version as I had decided to tackle the Showman's

Engine. This is proving to be a most satisfactory arrangement as Mark has

some very useful skills and has undertaken to obtain materials, it will be very useful to

have an extra pair of hands at boiler making time.

Examination of drawings obtained from Messrs. Plastow indicated that a number of

gears would need to be produced in a range from 13T to 96T in 10 DP plus bevel gears

and pinions and thought was given as to how these could be made. One of my most

used workshop tools is an almost complete set of 'Model Engineer's' and recourse was

had through the use of my self compiled master index to see how past contributors

had tackled the problem. In Model Engineer, 14 September 1950 an article by Base Circle describes a simple and elegant method of coupling a gear using a simple tool to generate teeth. As Base Circle points out this produces teeth of correct involute form whatever the

number being cut. Being the possessor of a small 'Perfecto' motorised shaper this

appeared to be the way to go, especially as my old WWI. Cater and Hakes No. 2

horizontal milling machine has insufficient capacity to handle the larger gears using

involute gear cutters. Fig. 1: Machiningthe blank for the 96 tooth gear on the miltingmachine.

Note the use of adaptors for the holder.

MODEL ENGINEER 20 JANUARY 1989

 

 

 

 

 

 

 

Having decided upon a method, the problem of producing the 96T blank had to

be solved. This blank is 9.8 in: dia. and I proposed to machine it from a manhole

cover casting. This, however was significantly larger than the 10 in. that my

Myford Super 7 lathe will swing in the gap. Another look at the milling machine showed it to be possible to machine a plug

mandrel for the spindle to accept Myford chucks and face plates, so this was turned

and Mark obtained a suitable casting upon which to mount the lathe topslide. Fig. 1 &

2 show the arrangement. Surprisingly for such a lash-up the rate of metal removal

and rigidity is quite good, and it was certainly very pleasant to have the power

cross feed available. A minor, but important detail was to dowel the casting, prior to machining the second face, to the faceplate for second operations, especially that of trepanning to an internal diameter after gearcutting. Incidentally, I am a firm believer in

production planning and a written procedure for such operations. This very largely

prevents the situation arising whereby you have removed metal now required for

location or measurement. In this instance I duced; it will be seen that the shaper table is

connected to the pitch circle disc by a wire tensioned between a "U" shaped metal bar

of substantial section affixed to the longitudinally moving portion of the machine.

Experience has proven that this wire must be very firmly anchored to the P.C.

Disc and be level; I used a line level. Initially indexing was done by an outer

ring of holes in the metal disc, which was indexed using the Myford dividing head. A

pin locates in a hole in the P.C. disc. When the nut at the rear is loosened, the blank is

rotated one hold and the nut tightened. ground upon the Quorn to dimensions

Right, Fig. 2: Machining the second face of the 96t. blank, Below, Fig. 3: Cutting the teeth in the 96t. gear, the final arrangement.Australian reader C. Bamford decided to

build traction engines. He chose to make all the spur and bevel gears first, using milling machine and shaper Part 1 obtained from Machinery's Handbook for a 10 D.P. rack. Following Base Circles suggestion a roughing and finishing cut was proposed GEARS attempted to foresee operations as a complete event to enable subsequent work such as keyway cutting to be carried out with automatic location upon a fixture. Thought had been given along these lines and cutting commenced. Immediately, as to how this relatively large casting could everything that was possible to shake loose be presented to the shaper tool, and did so, the crashing and banging was indexed, the final arrangement is shown in

horrendous and the amount of rocking at Fig. 3. the blank amazing.

A halt was called to tighten shaper gibs and packing placed under the angle plate

Method Adopted The basis is a large angle plate machined and between it and the bench, and a bar to slide vertically upon that portion of the

shaper that normally carries the work (fortuitously conveniently placed), to

holding table. The angle plate is accurately brace the blank. It was also found necessary machined on all four faces and the longer  to shim the clapper box to eliminate

face has so far proven to be the only part endwise movement and to pack the tool in

not possible to machine within my own the slot, I used a number drill for this.

workshop. Finally a small spring between clapper box To raise the blank to machining height

and slide was needed to return the box two further castings were obtained, one to

vertically after the return stroke. A grub carry a spindle which secures the blank at

screw replaced a gib screw in the vertical one end and the indexing arrangement at

slide to enable the slide to be firmly locked. the other. This casting was made to handle

Operations recommenced and eventually the remainder of the castings and thus a

by elimination, things settled down. supplementary raising block was required.

Fig. 3 also shows the connecting wire to the By fortunate chance I had the use of a pitch circle disc, the diameter of which is gear tooth vernier and this enable me to blank

 pcd minus wire diameter, in this case maintain a check upon tooth form. I was 0.050 inch. The figure shows the final amazed to find significant variation arrangement arrived at after some dearly between adjacent teeth. Following investibought experience, for whilst Base Circles gation it was discovered that slight varia method

works, his description is devoid of tions in the index hole spacing, probably

a number of pitfalls into which I fell. I caused by drill wander, was being mulsuspect

that he was cutting much smaller tiplied at the blank periphery and so the

gears than I was undertaking. method shown in Fig. 3 was substituted, 96

Fig. 1 of the original article is reproholes being drilled in the perimeter of the

MODEL ENGINEER 20 JANUARY 1989

 

 

 

 

 P.C. disc and a bar fastened to the index

plate, with the pin for indexing; this system proved adequate.

None the less, when the second, finishing, cut commenced it was immediately

clear that the required result was not being obtained and attention focussed upon the

tool. Investigation showed that the rack form tool tip was too broad and gradual slimming on an ad hoc basis eventually produced a suitable tool. Rightly or wrongly, I had calculated that to produce a rack the flattened tip must be 0.115 in. wide; however, in final form this reduced to 0.095 in. As two blanks were being machined together this meant that two further blanks had to be produced to replace the experimental pair, and this was done.

In case anyone is wondering why machining commenced with the most difficult

operation it was because we wished to complete the gears before ordering castings

from U.K. In the event of failure we had the option of availing ourselves of

Messrs. Plastow's gear cutting service. To shorten what could become a tedious

story, by the time the 13T gear was to be cut I had developed a method that was

giving very satisfactory results, in fact I was almost sorry to stop, Base Circle states

that it is an 'interesting' operation to watch, I found it fascinating. Operational Hints

For anyone considering using the method a few pointers may be of use:-

1). As mentioned, everything in the set-up must be as firm as possible. Obviously I

was overloading the machine with the 96 tooth blank, but even so a calamity ensued

on a smaller blank when the anchor screw on the P.C. Disc loosened. I found three

cuts per tooth gave best results for a depth of 0.20 in. (clearance at root of 0.015 in.), a

roughing cut all around of 0.015 in. and two at the final 0.20 in. depth. The first

finishing cut removed a majority of the amount which improves the finish and compensates for vertical movement of the blank under tool forces. 2). As the number of teeth reduce there appears to be a change of form but measurement proves that thickness at the p.c.d. remains constant. Maintaining this dimenstion effectively controls tooth form. 3). A problem not entirely overcome was that of measuring the tooth depth on the 13T gear due to the angularity of the teeth. However, by visual checking with the mating gear and maintaining the p.c.d. width a satisfactory result was obtained. Incidentally, by juggling with the tension

wire anchors it is possible to adjust the relationship between tool and tooth gap. Experience taught that the best results were obtained by ensuring that the relationship between blank and tool remained undisturbed. Top left. Fig. 4: The final arrangement and finishing cut on the 96t. gear, Above, Fig. 5: Fourteen gears, one set!Bevel gears and pinions in the foreground.Left, Fig. 6: Roughing the 21t. gear. Notereminder to set tool vertical. No. drill packing on the toot and the spring to returnthe clapper box. 4). The P.C. disc diameter must be accurate. I got some very funny shaped teeth when I forgot to subtract the tension wire diameter. I also found that no radius was required at the tool tip other than that required by normal tool grinding practice as the rotation of the blank formed an appropriate radius automatically. Tool wear was surprisingly little even when mild steel was being cut. The tool must be vertical. Fig. 5 shows the finished gears, together with the crankshaft and bevel gears and pinions. The method of producing these will be described next time, as will the

method of producing the various keyways. Fig. 6 is a close view of the 2IT gears

during the roughing cut whilst Fig. 7 shows the final cuts to the 13T gears. Apparent in

this view is the limited clearance for the tool at the 1.5 in. dia. required and the 2B A

bolt used to lock the spindle when indexing. In summary this has proven to be a

thoroughly practical method provided a shaper is available and a modicum of care

exercised. Admittedly one has to stay alert for problems. I would suggest, in order to

arrive at correct tool shape a little experimentation be undertaken, if possible upon

a somewhat smaller blank than those I used. A tool with a thinner tip is suggested

and this could be ground away as required until measurement of the tooth gives the

required result. Having said that I expect there is a simple method of arriving at the correct tool form, if so, I for one would be delighted to learn of it. To be concluded (Photocopies of "Base Circles" article can be obtained from our Reader's Services Dept. at this address - Ed). MODEL ENGINEER 20 JANUARY 1989

 

 

 

 

 

 

Part II (conclusion), from page 112 Part I of this mini-series told how the Author

adapted a method which was described in Model Engineer, 14 September 1950, to

produce gears for his Fowler traction engine. He described how he got over problems of

rigidity which arose. This time he goes on to discuss the methods adopted to produce the

spur and bevel gears for the model. He then talks about making keyways for various components. ith the spur gears successfully completed attention was

focussed upon the bevel gears and pinions for the compensating gear

(differential). Messrs Plastow's design calls for two 44T gears and two 9T pinions per

engine. He offers a specialised gear cutting service. However, having satisfied one minor ambition by producing the spur gears, I wished to pursue the project to completion by producing these more complex items. Back issues of the Model Engineer had provided a number of solutions, the most attractive being a method of cutting to a parallel depth. The method has advantages over conventional methods in that neither specialised machinery nor non-standard cutters are required. Mr. Minchin, who described the system in Model Engineer, 15 November 1964, claims that power transmission is 80% of conventional gears, but for my purposes (and I suspect most others) this small loss was of no moment. Again, this seemed to be a simple and elegant solution requiring only the use of a

horizontal mill. Machinery Used As mentioned in the previous article I

own an old Carter and Hakes No. 2 milling machine. This would seem originally to

have been belt driven from an overhead line shaft and is clearly a production

machine. Judging by certain wear it has had a hard life. It came into my possession

in the late 1960's for 5 and has been given a new mandrel, extended longitudinal

table traverse, chain drive and a graduated Australian reader C. Bamford decided to build traction engines. He chose to make all the spur and bevel gears first, using milling machine and shaper Above: Fig. 8 shows the set-up for pinions, tapered packing plate and blank holder. screw for vertical travel. An oddity is that both longitudinal table travel and vertical

feed can be traversed by a lever operated rack and pinion. Four speeds are obtainable, through what I understand, is a 1929 B.S.A. motor cycle gearbox. Reading and re-reading the article I decided upon a method somewhat different than that described by Mr. Minchin in that he used a dividing head to index the blanks, using vertical feed for the gears.

My milling machine has no power feed for vertical travel and so I was looking for a

method to utilise the table feed. Also, although owning a Myford dividing head, I

felt it might not be rigid enough for the very heavy work envisaged.

Having decided upon a method, I was fortunate in being able to contact Mr.

Minchin and he most generously amplified his article and provided drawings of blanks

and cutters. The cutters proved to be of quite simple shapes and he suggested silver steel, which in the event proved very satisfactory.  ig. 8 shows the basic set-up. A packing plate machined to a taper of 11 1/2 deg. to match the pinion and bevel angles is bolted square across the miller table. For the pinions the fixture previously used to

rotate the spur gear blanks was fastened to the packing plate and the blank bolted thereon.

Indexing Arrangement Fig. 9 shows the indexing arrangement.

An odd length of angle is bolted to the rear of the blank holder and a circle of plastic to

the shaft. Nine indexed holes are drilled The method calls for a central roughing cut and one each side with the cutter offset by 25% of the pitch with the blank rotated in the same direction also by 25% of the pitch. This was achieved by indexing the plastic direct to holes above and below the gashing index hole drilled through the angle. Spacing was taken from indexed holes drilled when dividing was carried out. Movement of the cutter was effected by

means of appropriate washers upon the mandrel. Cutting proved to be quite free in the

 

 

 

 

 

MODEL ENGINEER 17 FEBRUARY 1989

 

reversed for this blank and the table clamp adjusted to withstand climb milling forces.

Again, cutting was quick and clean. One of the most serious problems was to

locate the blank centrally beneath the cutter and this was achieved using square

and caliper. The appearance of teeth being offset in the figure is an optical illusion. Fig. 11 shows a set of completed gears, which were a pleasure to make, and which mesh perfectly, although by the time both sets (for two engines, remember) had been Above: Fig. 9 shows offset cutting. The cutter holder is offset by removing the washer, pinion rotated completed a certain tedium had set in. This and indexed to the top hole in the angle. Below: Fig. 10 shows climb milling the bossed bevel gear. The indexing details are apparent in this was dispelled by passing to the keywaying view, as is the simple tool shape of 20 deg. included. operations. Above right: Fig. 11, one set of completed bevel gears and pinions. cast iron used, at the slowest table teed. stub of metal to locate the gear bore with Keyway Cutting

Fig. 9 also shows the first offset cut. Both an extended stud, nut and washer for A number of the gears slide upon four washers are to the left of the tool holder clamping purposes. A countersunk screw splines and others are keyed to their shafts and the pin is indexing in the upper hole in and nut affixes the blank to an indexed disc by the usual single key, and a method was the bar. Also visible is the simple tool and holes, at suitable spacing, are drilled devised to use the shaper to form keyways holder and the indexing holes. into the packer. Fig. 10 shows the method For cutting the bevel gears the angled and the simple tool used. Because of the in the gears. The angle plate previously described was packing plate was modified to accept a boss in one gear, the table feed had to be set, short angle vertical, upon the Myford boring table and a 1.25 in. dia. hole bored through. A turned plug was indexed for four 1/8 in. dia. holes, clamped to the bored angle plate face and the holes transferred

to the plate. Gears requiring four keyways had a single hole drilled at the correct pcd and a

stub of 1/8 in. dia. metal pressed in lightly, locating in each of the four holes in turn

thus locating keyways accurately. Each gear to be keyed was arranged to have a 1.248 in. dia. spigot for locating into the bored hole, the spigot being removed as a final operation if not required. Bore diameters varied from 0.6 in. to 1 in. and key widths from 5/32 in. to 3/16 in. A suitable tool holder was affixed to the face of the vertical slide of the shaper which has

a convenient 5/8 in. diameter hole being utilised for the purpose.

Once the cutter was centralised, the table was locked and cutting proved quick and

easy, depthing was by using the slide's graduated collar. Unfortunately the

photograph taken of the operation met with a mishap, but the concept is simple

and, I am sure, well known to many. In conclusion, a most rewarding project

which gave me much pleasure and which taught me a lot, not least that the methods

most generously described by Base Circle and Mr. Minchin work, and work

well. My thanks to them both. MODEL ENGINEER 17 FEBRUARY 1989