patterns and castings are so intimately connected that it would be difficult to treat of them separately without continually confounding them together; it is therefore proposed to speak of pattern-making and moulding under one head.

every operation in a pattern-shop has reference to some operation in the foundry, and patterns considered separately from moulding operations would be incomprehensible to any but the skilled. next to designing and draughting, pattern-making is the most intellectual of what may be termed engineering processes—the department that must include an exercise of the greatest amount of personal judgment on the part of the workman, and at the same time demands a high grade of hand skill.

for other kinds of work there are drawings furnished, and the plans are dictated by the engineering department of machinery-building establishments, but pattern-makers make their own plans for constructing their work, and have even to reproduce the drawings of the fitting-shop to work from. nearly everything pertaining to patterns is left to be decided by the pattern-maker, who, from the same drawings, and through the exercise of his judgment alone, may make patterns that are durable and expensive, or temporary and cheap, as the probable extent of their use may determine.

the expense of patterns should be divided among and charged to the machines for which the patterns are employed, but there can be no constant rules for assessing or dividing this cost. a pattern may be employed but once, or it may be used for years; it is continually liable to be superseded by changes and improvements that cannot be predicted beforehand; and in preparing patterns, the question continually arises of how much ought to be expended on them—a matter that should be determined between the engineer and the pattern-maker, but is generally left to the pattern-maker alone, for the reason that but few mechanical engineers understand pattern-making so well as to dictate plans of construction.

[91]

to point out some of the leading points or conditions to be taken into account in pattern-making, and which must be understood in order to manage this department, i will refer to them in consecutive order.

first.—durability, plans of construction and cost, which all amount to the same thing. to determine this point, there is to be considered the amount of use that the patterns are likely to serve, whether they are for standard or special machines, and the quality of the castings so far as affected by the patterns. a first-class pattern, framed to withstand moisture and rapping, may cost twice as much as another that has the same outline, yet the cheaper pattern may answer almost as well to form a few moulds as an expensive one.

second.—the manner of moulding and its expense, so far as determined by the patterns, which may be parted so as to be 'rammed up' on fallow boards or a level floor, or the patterns may be solid, and have to be bedded, as it is termed; pieces on the top may be made loose, or fastened on so as to 'cope off;' patterns may be well finished so as to draw clean, or rough so that a mould may require a great deal of time to dress up after a pattern is removed.

third.—the soundness of such parts as are to be planed, bored, and turned in finishing; this is also a matter that is determined mainly by how the patterns are arranged, by which is the top and which the bottom or drag side, the manner of drawing, and provisions for avoiding dirt and slag.

fourth.—cores, where used, how vented, how supported in the mould, and i will add how made, because cores that are of an irregular form are often more expensive than external moulds, including the patterns. the expense of patterns is often greatly reduced, but is sometimes increased, by the use of cores, which may be employed to cheapen patterns, add to their durability, or to ensure sound castings.

fifth.—shrinkage; the allowance that has to be made for the contraction of castings in cooling, in other words, the difference between the size of a pattern and the size of the casting. this is a simple matter apparently, which may be provided for in allowing a certain amount of shrinkage in all directions, but when the inequalities of shrinkage both as to time and degree are taken into account, the allowance to be made becomes a problem of no little complication.

[92]

sixth.—inherent, or cooling strains, that may either spring and warp castings, or weaken them by maintained tension in certain parts—a condition that often requires a disposition of the metal quite different from what working strains demand.

seventh.—draught, the bevel or inclination on the sides of patterns to allow them to be withdrawn from the moulds without dragging or breaking the sand.

eighth.—rapping plates, draw plates, and lifting irons for drawing the patterns out of the moulds; fallow and match boards, with other details that are peculiar to patterns, and have no counterparts, neither in names nor uses, outside the foundry.

this makes a statement in brief of what comprehends a knowledge of pattern-making, and what must be understood not only by pattern-makers, but also by mechanical engineers who undertake to design machinery or manage its construction successfully.

as to the manner of cutting out or planing up the lumber for patterns, and the manner of framing them together, it is useless to devote space to the subject here; one hour's practical observation in a pattern-shop, and another hour spent in examining different kinds of patterns, is worth more to the apprentice than a whole volume written to explain how these last-named operations are performed. a pattern, unless finished with paint or opaque varnish, will show the manner in which the wood is disposed in framing the parts together.

i will now proceed to review these conditions or principles in pattern-making and casting in a more detailed way, furnishing as far as possible reasons for different modes of constructing patterns, and the various plans of moulding and casting.

in regard to the character or quality of wood patterns, they can be made, as already stated, at greater or less expense, and if necessary, capable of almost any degree of endurance. the writer has examined patterns which had been used more than two hundred times, and were apparently good for an equal amount of use. such patterns are expensive in their first cost, but are the cheapest in the end, if they are to be employed for a large number of castings. patterns for special pieces, or such as are to be used for a few times only, do not require to be strong nor expensive, yet with patterns, as with everything else pertaining to machinery, the safest plan is to err on the side of strength.

[93]

for pulleys, gear wheels, or other standard parts of machinery which are not likely to be modified or changed, iron patterns are preferable; patterns for gear wheels and pulleys, when made of wood, aside from their liability to spring and warp, cannot be made sufficiently strong to withstand foundry use; besides, the greatest accuracy that can be attained, even by metal patterns, is far from producing true castings, especially for tooth wheels. the more perfect patterns are, the less rapping is required in drawing them; and the less rapping done, the more perfect castings will be.

the most perfect castings for gear wheels and pulleys and other pieces which can be so moulded, are made by drawing the patterns through templates without rapping. these templates are simply plates of metal perforated so that the pattern can be forced through them by screws or levers, leaving the sand intact. such templates are expensive to begin with, because of the accurate fitting that is required, especially around the teeth of wheels, and the mechanism that is required in drawing the patterns, but when a large number of pieces are to be made from one pattern, such as gear wheels and pulleys, the saving of labour will soon pay for the templates and machinery required, to say nothing of the saving of metal, which often amounts to ten per cent., and the increased value of the castings because of their accuracy.

mr ransome of ipswich, england, where this system of template moulding originated, has invented a process of fitting templates for gear wheels and other kinds of casting by pouring melted white metal around to mould the fit instead of cutting it through the templates; this effects a great saving in expense, and answers in many cases quite as well as the old plan.

the expense of forming pattern-moulds may be considered as divided between the foundry and pattern-shop. what a pattern-maker saves a moulder may lose, and what a pattern-maker spends a moulder may save; in other words, there is a point beyond which saving expense in patterns is balanced by extra labour and waste in moulding—a fact that is not generally realised because of inaccurate records of both pattern and foundry work. what is lost or saved by judicious or careless management in the matter of patterns and moulding can only be known to those who are well skilled in both moulding and pattern-making. a moulder may cut all the fillets in a mould with a trowel; he may stop off, fill [94] up, and print in, to save pattern-work, but it is only expedient to do so when it costs very much less than to prepare proper patterns, because patching and cutting in moulds seldom improves them.

the reader may notice how everything pertaining to patterns and moulding resolves itself into a matter of judgment on the part of workmen, and how difficult it would be to apply general rules.

the arrangement of patterns with reference to having certain parts of castings solid and clean is an important matter, yet one that is comparatively easy to understand. supposing the iron in a mould to be in a melted state, and to contain, as it always must, loose sand and 'scruff,' and that the weight of the dirt is to melted iron as the weight of cork is to water, it is easy to see where this dirt would lodge, and where it would be found in the castings. the top of a mould or cope, as it is called, contains the dirt, while the bottom or drag side is generally clean and sound: the rule is to arrange patterns so that the surfaces to be finished will come on the bottom or drag side.

expedients to avoid dirt in such castings as are to be finished all over or on two sides are various. careful moulding to avoid loose sand and washing is the first requisite; sinking heads, that rise above the moulds, are commonly employed when castings are of a form which allows the dirt to collect at one point. moulds for sinking heads are formed by moulders as a rule, but are sometimes provided for by the patterns.

the quality of castings is governed by a great many things besides what have been named, such as the manner of gating or flowing the metal into the moulds, the temperature and quality of the iron, the temperature and character of the mould—things which any skilled foundryman will take pleasure in explaining in answer to courteous and proper questions.

cores are employed mainly for what may be termed the displacement of metal in moulds. there is no clear line of distinction between cores and moulds, as founding is now conducted; cores may be of green sand, and made to surround the exterior of a piece, as well as to make perforations or to form recesses within it. the term 'core,' in its technical sense, means dried moulds, as distinguished from green sand. wheels or other castings are said to be cast in cores when the moulds are made in pieces and dried. supporting and venting cores, and their expansion, are conditions to which especial attention [95] is called. when a core is surrounded with hot metal, it gives off, because of moisture and the burning of the 'wash,' a large amount of gas which must have free means of escape. in the arrangement of cores, therefore, attention must be had to some means of venting, which is generally attained by allowing them to project through the sides of the mould and communicate with the air outside.

an apprentice may get a clear idea of this venting process by inspecting tubular core barrels, such as are employed in moulding pipes or hollow columns, or by examining ordinary cores about a foundry. provision of some kind to 'carry off the vent,' as it is termed by moulders, will be found in every case. the venting of moulds is even more important than venting cores, because core vents only carry off gas generated within the core itself, while the gas from its exterior surface, and from the whole mould, has to find means of escaping rapidly from the flasks when the hot metal enters.

a learner will no doubt wonder why sand is used for moulding, instead of some more adhesive material like clay. if he is not too fastidious for the experiment, and will apply a lump of damp moulding sand to his mouth and blow his breath through the mass, the query will be solved. if it were not for the porous nature of sand-moulds they would be blown to pieces as soon as the hot metal entered them; not only because of the mechanical expansion of the gas, but often from explosion by combustion. gas jets from moulds, as may be seen at any time when castings are poured, will take fire and burn the same as illuminating gas.

if it were not for securing vent for gas, moulds could be made from plastic material so as to produce fine castings with clear sharp outlines.

the means of supporting cores must be devised, or at least understood, by pattern-makers; these supports consist of 'prints' and 'anchors.' prints are extensions of the cores, which project through the casting and extend into the sides of the mould, to be held by the sand or by the flask. the prints of cores have duplicates on the patterns, called core prints, which are, or should be, of a different colour from the patterns, so as to distinguish one from the other. the amount of surface required to support cores is dependent upon their weight, or rather upon their cubic contents, because the weight of a core is but a trifling matter [96] compared to its floating force when surrounded by melted metal. an apprentice in studying devices for supporting cores must remember that the main force required is to hold them down, and not to bear their weight. the floating force of a core is as the difference between its weight and that of a solid of metal of the same size—a matter moulders often forget to consider. it is often impossible, from the nature of castings, to have prints large enough to support the cores, and it is then effected by anchors, pieces of iron that stand like braces between the cores and the flasks or pieces of iron imbedded in the sand to receive the strain of the anchors.

in constructing patterns where it is optional whether to employ cores or not, and in preparing drawings for castings which may have either a ribbed or a cored section, it is nearly always best to employ cores. the usual estimate of the difference between the cost of moulding rib and cored sections, as well as of skeleton and cored patterns, is wrong. the expense of cores is often balanced by the advantage of having an 'open mould,' that is accessible for repairs or facing, and by the greater durability and convenience of the solid patterns. taking, for example, a column, or box frame for machinery, that might be made either with a rib or a cored section, it would at first thought seem that patterns for a cored casting would cost much more by reason of the core-boxes; but it must be remembered that in most patterns labour is the principal expense, and what is lost in the extra lumber required for a core-box or in making a solid pattern is in many cases more than represented in the greater amount of labour required to construct a rib pattern.

cores expand when heated, and require an allowance in their dimensions the reverse from patterns; this is especially the case when the cores are made upon iron frames. for cylindrical cores less than six inches diameter, or less than two feet long, expansion need not be taken into account by pattern-makers, but for large cores careful calculation is required. the expansion of cores is as the amount of heat imparted to them, and the amount of heat taken up is dependent upon the quantity of metal that may surround the core and its conducting power.

shrinkage, or the contraction of castings in cooling, is provided for by adding from one-tenth to one-eighth of an inch to each foot in the dimensions of patterns. this is a simple matter, and is accomplished by employing a shrink rule in laying down pattern-drawings [97] from the figured dimensions of the finished work; such rules are about one-hundredth part longer than the standard scale.

this matter of shrinkage is indeed the only condition in pattern-making which is governed by anything near a constant rule, and even shrinkage requires sometimes to be varied to suit special cases. for small patterns whose dimensions do not exceed one foot in any direction, rapping will generally make up for shrinkage, and no allowance is required in the patterns, but pattern-makers are so partial to the rule of shrinkage, as the only constant one in their work, that they are averse to admitting exceptions, and usually keep to the shrink rule for all pieces, whether large or small.

inherent or cooling strains in castings is much more intricate than shrinkage: it is, in fact, one of the most uncertain and obscure matters that pattern-makers and moulders have to contend with. inherent strains may weaken castings, or cause them to break while cooling, or sometimes even after they are finished; and in many kinds of works such strains must be carefully guarded against, both in the preparation of designs and the arrangement of patterns, especially for wheels and pulleys with spokes, and for struts or braces with both ends fixed. the main difficulty resulting from cooling strains, however, is that of castings being warped and sprung; this difficulty is continually present in the foundry and machine-shop, and there is perhaps no problem in the whole range of mechanical manipulation of which there exists more diversity of opinion and practice than of means to prevent the springing of castings. this being the case, an apprentice can hardly hope for much information here. there is no doubt of springing and strains in castings being the result of constant causes that might be fully understood if it were not for the ever-changing conditions which exist in casting, both as to the form of pieces, the temperature and quality of metal, mode of cooling, and so on.

castings are of course sprung by the action of unequal strains, caused by one part cooling or 'setting' sooner than another. that far all is clear, but the next step takes us into the dark. what are the various conditions which induce irregular cooling, and how is it to be avoided?

irregularity of cooling may be the result of unequal conducting power in different parts of a mould or cores, or it may be [98] from the varying dimensions of the castings, which contain heat as their thickness, and give it off in the same ratio. as a rule, the drag or bottom side of a casting cools first, especially if a mould rests on the ground, and there is not much sand between the castings and the earth; this is a common cause of unequal cooling, especially in large flat pieces. air being a bad conductor of heat, and the sand usually thin on the cope or top side, the result is that the top of moulds remain quite hot, while at the bottom the earth, being a good conductor, carries off the heat and cools that side first, so that the iron 'sets' first on the bottom, afterwards cooling and contracting on the top, so that castings are warped and left with inherent strains.

these are but a few of many influences which tend to irregular cooling, and are described with a view of giving a clue from which other causes may be traced out. the want of uniformity in sections which tends to irregular cooling can often be avoided without much loss by a disposition of the metal with reference to cooling strains. this, so far as the extra metal required to give uniformity to or to balance the different sides of a casting, is a waste which engineers are sometimes loth to consent to, and often neglect in designs for moulded parts; yet, as before said, the difficulty of irregular cooling can in a great degree be counteracted by a proper distribution of the metal, without wasting, if the matter is properly understood. no one is prepared to make designs for castings who has not studied the subject of cooling strains as thoroughly as possible, from practical examples as well as by theoretical deductions.

draught, or the taper required to allow patterns to be drawn readily, is another of those indefinite conditions in pattern-making that must be constantly decided by judgment and experience. it is not uncommon to find rules for the draught of patterns laid down in books, but it would be difficult to find such rules applied. the draught may be one-sixteenth of an inch to each foot of depth, or it may be one inch to a foot of depth, or there may be no draught whatever. any rule, considered aside from specified conditions, will only confuse a learner. the only plan to understand the proper amount of draught for patterns is to study the matter in connection with patterns and foundry operations.

patterns that are deep, and for castings that require to be parallel or square when finished, are made with the least possible amount of draught. if a pattern is a plain form, that affords [99] facilities for lifting or drawing, it may be drawn without taper if its sides are smooth and well finished. pieces that are shallow and moulded often should, as a matter of convenience, have as much taper as possible; and as the quantity of draught can be as the depth of a pattern, we frequently see them made with a taper that exceeds one inch to the foot of depth.

moulders generally rap patterns as much as they will stand, often more than they will stand; and in providing for draught it is necessary to take these customs into account. there is no use in making provision to save rapping unless the rapping is to be omitted.

rapping plates, draw-irons, and other details of pattern-making are soon understood by observation. perhaps the most useful suggestion which can be given in reference to draw-irons is to say they should be set on the under or bottom side of patterns, instead of on the top, where they are generally placed. a draw-plate set in this way, with a hole bored through the pattern so as to insert draw-irons from the top, cannot pull off, which it is apt to do if set on the top side. every pattern no matter how small, should be ironed, unless it is some trifling piece, with dowel-pins, draw and rapping plates. if a system of draw-irons is not rigidly carried out, moulders will not trouble themselves to take care of patterns.

in conclusion, i will say on the subject of patterns and castings, that a learner must depend mainly upon what he can see and what is explained to him in the pattern-shop and foundry. he need never fear an uncivil answer to a proper question, applied at the right time and place. mechanics who have enough knowledge to give useful information of their business, have invariably the courtesy and good sense to impart such information to those who require it.

an apprentice should never ask questions about simple and obvious matters, or about such things as he can easily learn by his own efforts. the more difficult a question is, the more pleasure a skilled man will take in answering it. in short, a learner should carefully consider questions before asking them. a good plan is to write them down, and when information is wanted about casting, never go to a foundry to interrupt a manager or moulder at melting time, nor in the morning, when no one wants to be annoyed with questions.

[100]

i will, in connection with this subject of patterns and castings, suggest a plan of learning especially applicable in such cases, that of adopting a habit of imagining the manner of moulding, and the kind of pattern used in producing each casting that comes under notice. such a habit becomes easy and natural in a short time, and is a sure means of acquiring an extended knowledge of patterns and moulding.

a pattern-maker no sooner sees a casting than he imagines the kind of pattern employed in moulding it; a moulder will imagine the plan of moulding and casting a piece; while an engineer will criticise the arrangement, proportions, adaptation, and general design, and if skilled, as he ought to be, will also detect at a glance any useless expense in patterns or moulding.

(1.) why cannot the regular working drawings of a machine be employed to construct patterns by?—(2.) what should determine the quality or durability of patterns?—(3.) how can the arrangement of patterns affect certain parts of a casting?—(4.) what means can be employed to avoid inherent strain in castings?—(5.) why is the top of a casting less sound than the bottom or drag side?—(6.) what are cores employed for?—(7.) what is meant by venting a mould?—(8.) explain the difference between green and dry sand mouldings.—(9.) why is sand employed for moulds?—(10.) what generally causes the disarrangement of cores in casting?—(11.) why are castings often sprung or crooked?—(12.) what should determine the amount of draught given to patterns?—(13.) what are the means generally adopted to avoid cooling strains in castings?