By JO H N HIRD (Associate Member) This P aper is a description of the m aking in
green sand o f a 4 ft. 6 in. dia. electric m otor casing or shell weighing 22 cwts. and the end covers for the same m otor. W hen instructions were received to m ake this particular casting, the largest casting so far m ade in the foundry was 2 ft. 4 in. dia.
Deciding Upon the Moulding Method The usual m ethod of m aking these castings is in dry sand or by striking them up in loam,
Fi g. I . — Co m p l e t e d Ca s t i n g Re a d y f o r Ma c h i n e Sh o p.
thus avoiding the need of making a complete pattern. F o u r factors had to be considered when deciding upon the m ethod to be a d o p te d : (1) The crane is only o f 3 tons capacity; (2) there is no suitable m ould-drying stove; (3) the foundry floor is concreted all over; and (4) the quantities off these castings, while not initially large, were likely to increase in the future.
It was decided th a t if the m oulding boxes were kept as light as possible and a minim um am ount o f sand used, the crane would be adequate. This was possible if the green-sand m ethod was used, and it eliminated the second and third factors. T he m aking of a perm anent pattern suitable for green-sand m oulding satis
fied the fourth factor. A further consideration was that, should it prove im practicable to make
these castings in green sand, the cost of a waster would be small com pared with that o f a core stove sufficiently large to dry these moulds.
There is, of course, the alternative and “ half
way house,” skin drying, which is often u n satisfactory owing to uneven drying, and the risk o f burning in one p art while other parts rem ain green or strike back. On these con
siderations it was decided to m ake the first casting in green sand.
Fig. 1 shows the com pleted shell casting ready to go to the m achine shop. It is 4 ft.
6 in. dia. and 2 ft. 4 in. wide, with a wall thick
ness of J in. There are ten ribs, 2 \ in. thick by 6 in. wide in the centre, and stepped down at each end to 3 in. wide for a distance which varies at each end. The flange round each end is 1 in. thick and 2} in. wide.
The triangular holes in the feet are at both sides o f the shell and were asked for by the
Fi g. 2 .— We l d e d Bo x Pa r t s Us e d.
foundry to help to su pport the feet cores and enable the feet core prints to be kept small, and so allow the size of the m oulding box to be a t a m inim um . A ctually the feet core prints are only 1-J- in. wide. This is a good exam ple of the excellent co-operation which exists be
tween the foundry and the draw ing office. As every foundrym an knows, such co-operation is
Fi g. 3 .— Ha l f Pa t t e r n f o r Mo t o r Ca s i n g.
a great asset in m aking satisfactory castings economically.
T he hole in the side of the shell is for lead
ing the cables to a term inal board. This feature is not perm anent, but will be altered to suit the wiring needs o f each order.
M oulding Boxes
T he welded steel m oulding boxes, shown in Fig. 2, were designed and purchased, the m am reason for using such boxes being to keep the w eight within the capacity of the crane. The total weight of these boxes, w hich are 5 ft.
square, is 27 cwts. The to p box alone weighed 10 cwts. T he dim ensions left only 3 in. of sand between the edge of the pattern and the box.
One top part was m ade 17 in. deep, and the three other parts to be used as middle or botlc»-Ti parts as required were 4 in., 6 in. and 12 in.
deep. T he various depths are arranged so that the boxes can be built up to suitable depths for m aking the end-cover castings, o f which there are two depths. T he bars in the top box are arranged so that there is a 2-ft. square opening in the middle, fitted with detachable bars to be taken out when m aking the shell casting. This is to allow a 2 ft. square by 1 ft. deep exten
sion box to be added when m aking sleeve- bearing end-covers.
Lugs are arranged on all parts, so that any assembly can be bolted together from top to bottom by fo u r 1 in. dia. bolts, and 1| in. dia.
pins are used for locating the pattern and
m oulding boxes. A f-in. thick bottom plate was purchased with a 2-ft. dia. hole in the centre and strongly ribbed on the back, the fro n t being m achined all over, as were all joint faces on the m oulding boxes.
Shell P attern Em ployed
A half-pattern (Fig. 3) was m ade and mounted centrally on the pattern board. T he pin-holes are placed between the feet. Both top and bottom halves o f the m ould are ram m ed off this half-pattern. The print to take a core to form the hole in the side can be changed over from one side o f the pattern to the other, for top and bottom moulds.
As several widths o f shell will be required, the pattern is m ade to the narrow est width that will be w anted, and the print for the body core is m ade 3 | in. long. This leaves a H -in. long print when the widest shell is being made.
Rings fit over this print to m ake the pattern width up to th a t required for any particular shell casting. This can vary from 28 to 32 in.
The depth of the m oulding box is arranged so that the print comes to the top o f the box in both to p and bottom moulds. T here being no sand over the print, this is possible, as the height o f the print does not vary, whichever width o f shell is being made. This m akes a hole right through the mould.
In the bottom half the 2-ft. dia. hole in the bottom plate is a few inches sm aller than the diam eter o f the print. In the top half-m ould
Fi g. 4 . — Ha l f Co r e Bo x e s Us e d.
the 2-ft. square section fitted with the loose bars has the bars rem oved, and the print comes up to the bottom edge of the bars. N o sand is ram m ed in the square opening. R ound rods, 1| in. dia., are placed on the feet in the centre of the place w here the triangular pad on the feet cores cuts through. These rods are wedged under the bars in the top part, and rest on the 138
bottom plate in the bottom part. They con
nect directly on to other rods running right through each half of the foot cores, and ensure that the foot cores will not lift when casting.
The 1 | in. wide prints on the feet only act as location prints, and are not intended to take much of the weight or lift of the cores.
Core Boxes
A half core box (Fig. 4) was made, as the two halves of the core are identical, with the exception of the prin t for the round cored hole in the side, which is changed over for the top and bottom half-cores. T he body o f the core is m ade to suit the widest shell required. The bottom of the print is m ade to the same depth from the joint as the top o f the print on the pattern to the joint.
Rings are placed in the bottom of the body o f the core box when no t m aking the widest shell, to bring the body of the core down to
the width required. Conversely, the rings are taken off the pattern when not m aking the widest shell, to bring the height of the pattern down to the width required. The 2 |-in . by 6-in.
ribs previously m entioned as being stepped down at each end are the full w idth right through the core box, the step being m ade by inserting green-sand cores o f the required length into the recess form ed by the ribs on the body core, before coring-up. The steps are of a different length at each end of the shell, and vary in different shells from 2 to 6 in. at each end.
These cores eliminated the necessity for inter
changeable loose ribs in the core box. As there are at least six varying lengths o f these ribs, this obviated a considerable am ount o f pattern- making.
Original Design of Core Grid
The grid shown in Fig. 5 is a ring, open at both ends, m ade of |-in . plate with \ in. dia.
F i a . 5 .— F i r s t T y p e o f C o r e G r i d . 139
holes at 3-in. centres, stiffened by two 1-in. dia.
rods across each end. G aggers, m ade by threading 3-in . dia. rods 1 j in. a t one end and locking through the holes in the grid with a nu t each side, were used. These were placed so as to be each side o f the ribs in the core box, and were extended to w ithin 1 in. o f the
the hole in the bottom plate, and straps on to the 1-in. dia. stiffening bars in the grid.
The to p half of the core, shown ram m ed up in Fig. 6, is turned over on the trunnions after a round board has been bolted on to the top of the core box, right through the grid, and the bottom o f the core box. T he bolts are then released and the core box lifted off with the slings through fo u r 1-in. eye-bolts screwed into nuts, let into the bottom o f the core box. The core is then picked u p with the eye-bolts in the grid and placed on to p o f the bottom core.
The rims o f the grids m ake a m etal to m etal contact, and the to p core is bolted down to the bottom core with straps across the 1-in. dia.
bars in the grids.
T he bottom -half foot cores are pu t in the mould, and the top half placed on them , with oil and blacking on the joint. All the air is brought off from the bottom core. These cores and the round one for the hole in the side are dried. They are m ade from o rdinary backing sand, bonded w ith 2 per cent, bentonite.
Fi g. 6 .— To p Ha l f o f t h e Co r e.
outside of the core box. A bout 120 of these were used in each half of the core. It was thought that, being a green-sand core, the sand would give when the contraction cam e into play, and th a t provision need no t be m ade for contraction in designing the grid.
T he core, whilst being 4 ft. 2 in. across, has only 2 in. o f sand from the root of the ribs to the grid, which was m ade to fit in the print at the bottom of the core box, and was exactly level with the joint. T he object of the grid, besides carrying the sand, is to vent the core quickly and freely, and to minimise the weight of the core. T he bottom half-core is lifted out of the core box w ith an adjustable four-hook sling. This is assisted by using the ring in the Dottom o f the core box as a lifting or stripping plate. F o u r i-in . rods, threaded at each end, are arranged w ith a nut, countersunk under the
;ing. Bars threaded over the top of the rods are bolted to the grid. T he core when lifted out, and the bottom ring rem oved by taking off the nuts under the ring, will stand on a board, as the weight is taken on the rim o f the grid.
T he grid in the bottom core rests over the 2-ft. dia. hole in the bottom plate, and the bottom rim o f the grid m akes a m etal to m etal contact with the bottom plate. W hen the core is pu t in the m ould, it is bolted to the bottom Dlate by m eans o f a piece of channel iron across
Fi g. 7 . — Ru n n e r Ar r a n g e m e n t.
140
S E C T I O N
Running Arrangements
These castings have to be run at several points on the diam eter, as it has been found, with m uch smaller shells, that if they are run a t one point the castings are not symmetrical, owing to all the hot m etal being at one side o f the m ould; the cooling is then uneven. This shell is run as is shown in Fig. 7 straight off the top w ith fo u r f-in. dia. dow n-runners direct on to the ribs. Three 11-in. dia. risers are brought off the ribs between the runners.
Stoppers are used until the bush is full, when all fo u r are lifted simultaneously.
The first m ould was made, closed and cast.
Everything went according to plan, and it was thought that all was well. H owever, next m orn
ing during stripping a | in. wide crack was
found across the width of the shell, through the round hole in the side. It was realised that, as the thickness o f sand in the core is com paratively small com pared with the weight of iron, particularly in the ribs, the sand had baked hard, like a dry-sand core. This was also assisted by the good venting, and baking had taken place before the contraction set in. It was afterw ards learnt that about 4 a.m. the following m orning the m en on duty at the A.F.S. station had heard two loud reports which they could no t account for. It would seem that the otherwise good casting broke through one side o f the hole first, followed very shortly afterw ards by the other side.
A nother casting was started as soon as the core grids had been modified to allow for
con-P i o . 8 .— Se c o n d Ty p e o f Co r e Gr i d. 141
traction. T here was also an o th er fa u lt in the grids which becam e ap p a ren t w hen the core was rem oved. T he sand was difficult to get at, as the grid extended the full w idth o f the casting. It took three men a day to rem ove the core.
Core Grid Modified
In m odifying the grid, as show n in Fig. 8, this was kept in m ind. Instead o f splitting the ring and taking ab o u t 2 in. ou t at one point an d strapping it, the grid was out into four pieces by drilling -¿-in. dia. holes across its w idth and grinding off the jagged edges. This left a \ in. gap at each joint, 2 in. in all, to allow for initial contraction. T he segments are held together by two cast-iron rings in each half-grid, and 1-in. threaded rods are screwed into rings, two to each segment. O thers are fastened in the segm ent w ith nuts each side, exactly opposite the rods in the rings. These rods are fastened together by screwed sleeves, thus holding the whole grid together. T he top and bottom rings in each half-grid are held in position by four vertical rods f-in . dia.
T he sam e i-in . dia. threaded rods were used for gaggers.
In this second shell, the sleeves w ere screwed back after casting and the nuts taken off the
Fi g. 9 . — Fi n a l Mo d i f i c a t i o n o f Co r e Gr i d.
vertical rods. T hen the centre rings were lifted out. The inside nuts were taken off the gagger rods and this allowed the segments to be draw n inw ard and lifted out. T he casting was now free to contract and it was easy to knock out the core.
It was decided to rem ove two opposite seg
m ents in eaoh grid im m ediately after casting.
In theory this m ay seem easy, but in practice it was an exceedingly unpleasant job, working
in the centre o f the m ould. I t required four m en from 4.30 (the casting was poured a t 4.0) until 9.0 to do the jo b w orking in relays. T his casting was sound, and as the second one o f the order h a d still to be m ade, a fu rth er m odifica
tion to the grid was m ade, in order to ease the w ork o f releasing it after casting.
Final Arrangement of Core Grid T he rings were now turned to carry a taper and w ere cut into fo u r pieces, and a com ponent
Fi g. 1 0 .— En d Co v e r Ca s t i n g.
was chuck-turned to the sam e taper to receive them. T he w hole assem bly was held together w ith two clam ps in the form o f a cross with a 1-in. bolt through the centre as show n in Fig. 9. T he threaded rods an d sleeves were used to hold the grid segments to the ring seg
m ents, as they were already in existence, but the sleeves were no t screwed ou t after casting, as the grid now cam e into sections w hen the centre bolts w ere released.
In place o f the y-in. dia. gagger rods, cast- iro n gaggers were m ade. These consisted o f a single casting carrying six gaggers, the back of w hich was drilled and tapped and held in posi
tion on the grid w ith a f-in . bolt a t each end.
T he fro n t h ad six gaggers \ in. by \ in. project
ing out 1 \ in. Tw o o f these were arranged between each rib, m aking a total of 20 on each grid. T he unpleasant job o f releasing the core was m uch easier and was reduced from 4± to 2 hrs. This casting was, in the w ords
o f the standard specification for iron castings, sound, clean, out o f twist, and free from blow
holes, distortion and all surface defects. The casting was m ade from start to finish by two men in three days.
The End-Covers
T here are two end-cover patterns for these machines, one deep and the other shallow. The variable speed a.c. com m utator m otor uses one shallow and one deep cover. O ther types take two shallow covers. Each type of m achine is
Fi g. 1 1 .— Pa t t e r n a n d Gr i d f o r t h e En d Co v e r.
m ade with either sleeve bearings or ball bear
ings, which involves a change-over o f bearing bosses for each end-cover to avoid m aking four patterns.
Fig. 10 shows the shallow end-cover casting carrying a sleeve bearing. The general thicknes;
of this casting is f in. There are six windows for ventilating, two on each of the three sides.
The baffle round the inside is 1 in. thick and projects inwards a m ean distance of 4 | in. The diam eter of the casting is 4 ft. 6 in., the same as the shell. The depth is 13 in. and the weight 8 cwts.
Fig. 11 shows the pattern and grid for the shallow cover. A solid block pattern was built up for the outside o f the pattern, and m ounted on a pattern board. The inside was built up, and fastened into a fram e, which form ed the pattern board. T he windows and baffle are form ed by four loose pieces which extend up to the joint on the pattern board.
The pattern is fitted up for m aking the plain or sleeve-bearing type of cover. The ball-bear
ing boss is leaning against the fro n t of the pattern. The ball-bearing cover will be dealt with next.
M oulding the Ball-Bearing Cover The m oulding boxes are those used for the shell casting split up into suitable depths. In this case, the 6-in. middle part is used for the
bottom part. The 4-in. middle part is bolted to the 17-in. top p a rt which has the loose bars in the centre inserted, and form s the top part.
The inside of the cover is m oulded bottom downwards. The grid shown on the right of
The inside of the cover is m oulded bottom downwards. The grid shown on the right of