Foundry practice; a treatise on molding and casting in their various details

Introduction

In administering the work in foundry practice at the University of Minnesota, the want of a good text book has been a serious disadvantage. The work of the shop and that of the class room should be correlated — shop work should be studied and discussed in the class room, and examples illustrating the various principles underlying good practice should be worked out in the shop.

While there have been some excellent books written upon the subject of foundry practice, yet, as a rule, these have been written with the needs of the experienced molder in view rather than those of the beginner. For this reason it is a difficult matter to teach the subject so that the student will acquire an intelligent understanding of its various details. The nomenclature and shop phraseology are not sufficiently elementary for the average beginner to grasp the statement presented, and much time is frequently spent in explaining an author's meaning.

The present little treatise has been written with a full knowledge of the problems involved and with the object of lessening some of the difficulties which arise in teaching tJ"' subject. The authors are both men of wide experience in foundry practice and its correlated subjects. Air. Tate is an experienced pattern maker, who has been in charge of the pattern shop at the University of Minnesota for the past fifteen years, and during a part of this time he has also had charge of the work in the foundry. Mr. Stone is a graduate of the University, who has given especial attention to foundry work, both from the iv INTRODUCTION

standpoint of the chemist and from that of the molder.

In piesenting this work on foinidry practice, the authors reahze that it is not a complete treatise on the subject. The aim has been to produce a book in which the principles of foundry practice are set forth concisely and clearly ; with the needs of the engineering student in view rather than those of the practical foundryman. To this end numerous examples are given representative of the different kinds of molding, and it is believed that the simple methods used in illustrating and describing the various operations involved and the reasons therefor will give the student a ready knowledge of the details of molding which will go far to supplement the practical work of the foundry, which, in a college course, must necessarily be limited.

While the treatment is thus somewhat brief, the subject matter as here presented is intended to cover all ordinary work in foundry practice including both brass and iron casting.

A glossary of foundry terms has been added, as it has been found that to obtain the greatest value from a work "of this character, the reader must become familiar with names and expressions used by foundrymen, for even if it were possible to eliminate shop expressions,

it would be undesirable to do so.

J. J. Flather.

Professor of Mechanical E7igi?teerino,

Mififieapo/is, Miniiesota. University of Minnesota.

Septejnber, igo^.

The authors wish to acknowledge their indebtedness to Mr. E. A. Johnson, Instructor in Foundry Practice at the University of Minnesota, and also to other foundrymen for information and suggestions received in the preparation of their work.

Chapter I

A sand suitable for molding must be open to allow the escape of gases and must be able to hold a given form to withstand pressure and wash of the metal. Such a sand has a percentage of clay or binding material which will hold the mass together firmly when dampened and compressed. If the percentage of clay becomes too great, the sand is too close when compressed, SO' the gases cannot pass off ; then the metal will not lie quietly against the face of the sand.

The molding sands used in diiTerent parts of the country vary greatly in their composition. Those high ni clay must be used with as little water as possible and must not be compressed or rammed much as the mold must give free escape for gases through the sand. The coarse sands very low in clav may require much water and hard ramming in order to form a satisfactorv mold. The tempering and ramming of the sand must be largely gauged by the nature of the sand the molder has at hand.

Tempering the sand means the mixing and wetting of the sand ready for making a mold. It is otherwise known as cutting over the sand.

The sand should be mixed evenly and to a dampness such that it will stick together when squeezed in the hand, but not so wet as to show moisture or dampen the hand. The sand pile should be opened out so that there will be no holes in which the water will accumulate. The water should then be thrown over the sand ni thin sheets by swinging the pail with the bottom slightly ahead of the top. In this manner the water is distributed evenly and does not cause mud in spots. If the sand is wet excessively in spots as by throwing the water on the pile in a body, it requires much more shovelling to obtain an even temper, hence loss of time. The sand should then be shovelled over in order to mix thoroughl}-. The shovelling should be done so^ as to scatter the sand when casting it from the shovel. This is accomplished by giving the handle of the shovel a twist just as the sand is leaving it. When wishing to throw the sand to a distant point, it should be allowed to leave the shovel in a solid mass, but this does not mix it evenly. In mixing, a space should always be kept between the pile from which the sand is taken and the one to which it is thrown. If this is not observed some of the sand will not be thoroughly mixed. After the sand has been shovelled over once it seldom is found to be mixed thoroughly, which makes it preferable to cut it over from two to three times. All the water necessary for the proper tempering should be put on before shovelling over the sand the last time. When trying to find whether the sand needs more water or not the hand should be forced into the pile to get some sand from the interior from which to determine its temper. This should be done at several points. The sand from the interior from which to determine its that in the heap. When only a little more water is necessary it should be sprinkled on by throwing the water from the pail with the hand.

The molder or helper should learn to shovel either right- or left-handed, so as to be able to take either side of the heap when working with an assistant.

The riddle is the sieve used for sifting the sand. Its meshes range from 2 to the inch to 16 or 32 per inch. They are numbered according to the number of meshes per inch, as a No. 2 riddle means one having y2 in. meshes, a No. 4 has Yx in. meshes, a No-. 16 has ViG ii'i- meshes, etc. In some places the riddles having the mesh finer than ^ in. are called sieves.

In riddling sand by hand, the riddle should be held loosely in the hand and carried by the fingers so that the palm of the hand will strike the rim as it is cast from side to side. Hitting the rim of the riddle in this way jars loose the sand that sticks to the riddle, keeps the meshes open better, and allows the sand to pass through more freely. By practice in holding the riddle in this manner, a rocking swing may be obtained which jars the riddle at each turn and carries but very little weight on the fingers. It is often found of advantage, especially in fine riddles, to put some irons in with the sand, as gaggers, etc. These irons scrape the wires clean and add to the jarring of the riddle.

When not in use, the riddle should always be hung up on a nail or placed on the sand heap with the screen up. If left with the screen resting on the sand, the meshes become clogged, thus hindering the passage of the sand through the screen.

There are many forms of mechanical sand sifters. The two representative forms of pneumatic sifters are shown in Figs. 93 and 96, while the belt-driven sifters are shown in Figs. 98 and 99.

Facing sand is placed next to the pattern in making a mold in order that the sand will peel or part from the casting freely and leave a smooth surface. Facing sand contains a percentage of sea coal and usually new sand, dependent upon the kind of work for which it is to be used.

The percentage of sea coal varies greatly, depending upon the thickness of metal and type of casting. The limits are i part of sea coal to 2 parts of sand, and i part of sea coal to 16 to 20 parts of sand. The limiting proportions are very seldom used. The usual proportions are from i to 6, to i to 14 of sand, depending on the thickness of the metal. When the metal is thinner than ^2 in. no facing is necessary. Better and smoother castings are obtained in this case by using heap sand riddled through a fine riddle onto the pattern. For metal between y'l in. and i in. the proportion should be about i part of sea coal to 12 or 14 parts of sand; between i in. and 2 in., i part of sea coal to 8 or 10 parts of sand; above 2 in., i part of sea coal to 6 or 8 parts of sand.

The sand used in the facing may also vary in its proportion of new and old sand. This is dependent upon the sand used. The most general proportion is 1 part of new sand to from 3 to^ ■; parts of old sand. Greater percentages of new^ sand mav h^ used on heavy work. The limiting case is a facing made of entirely new sand for the cope of very heavy work.

It is not always the thickness of the casting that regulates the strength of the facing sand. There are many other things to be considered: (i) whether the casting is to be poured with hot or dull iron; (2) the distance of some parts of the mold from the gate; (3) the time it will take the mold to become filled with iron ;

(4) whether the metal is running over flat surfaces, and

(5) is covering them slowly or quickly. Then again, heavv solid castings have become "cold-shot'' owing to the use of facings that were weak in proportion to the casting, caused by the slow rising of the metal in pouring. Strong facings on the sides oi a mold, where the iron enters and rises slowly, may easily cause heavv castings to be ''cold-shot." Again, the square corners of castings should, in general, have weaker facings than the straight, plain surfaces. The lower parts of deep molds should have a stronger facing than the upper portion, because the metal becomes dull while rising to the top of the mold. If the facing suitable for the lower portion were used at the upper, the casting at the upper part would become curly or partly cold-shot at the surface. A new sand without mixture will require more sea coal than if it were mixed with old or common heap sand.

A thorough mixing of the facing is necessary. If the sea coal is not evenly mixed, it often causes the casting to be streaked, veined, or cold-shot.

In mixing by hand it is almost impossible to distribute the sea coal evenly, therefore it is important that it should be handled several times in order to come as near as possible to a thorough mixture.

In mixing, the old and new sand should be kept as dry as possible when shovelled over in order to mix well. The sea coal is added while the sand is spread out thin. The whole is cut over once or twice, then riddled through a No. 6 or 8 riddle. It is then tramped down and water put on to^ give the proper temper, as in the case of tempering the heap s-md. It is again cut over to mix the wet and dry sand, then riddled through a No. 4 riddle. It is now ready to be riddled onto the pattern. The mixture should always be riddled twice, and better still, three or four times It is best to use sand quite dry to start the mixture, as when wet the sea coal sticks in small balls and does not mix well. In large foundries, the facing sand is mixed by a

Fig. I.

facing machine which gives a mixture of exact proportions and more thoroughly mixed than can be done by hand.

The frame in which a mold is made is called a flask. It is composed of two or more parts. The bottom part is trailed the drag or nowel, the top part is called the cope, and the intermediate parts, when used, are called the cheek. Plasks are made of wood or iron.

The form of flask used for small patterns when the pressure of the metal is very little, is represented in Fig. I. These are called snap flasks. They are hinged at one corner and fasten at the diagonal corner with a snap. The mold is rammed in the flask and when ready for pouring the flask is unsnapped and removed. Thus many molds may be made with a single flask. Before casting, a frame the same size as the flask is placed around the body of sand and a weight is placed on top to prevent straining the mold when under pressure.

Small flasks up to^ 14 in. square may best be made of iron, without bars in the cope. Those larger maj, be of either wood or iron to suit the style of work to be put in the flask. When the flask is for a special pattern and is to be used for that only, an iron flask will give the better service and is far cheaper. When for general patterns, the wood flask has many advantages. The bars may be fitted to a pattern in wood with little time or expense, and for a class of small work up to 40 in. square, may be made of suflicient strength. Larger flasks are made of wood having bolts or iron bars for stiffening the cope. When but a single casting is desired, even to very large castings, the flask mav be more cheaply made with a wood frame and iron bars than entirely of iron.

In manufacturing shops having a fixed line of patterns, the iron flask is of great value. The first cost is more than that of a wood flask, but the durability far exceeds that of wood. The bars are shaped to suit the pattern, and they remain so ; while a wood bar burns out and the joint of the wood flask burns away leaving holes which may cause a run-out, thus losing the casting. The iron flask is much heavier to handle, but it mav be fitted so as to require less anchoring in the sand, as gaggers and soldiers which would be required in a wood flask; thus the time saved in molding will more than ecjual the extra help necessary to handle the flask. Large flasks to hold castings as cylinders, engine girders, bed castings, etc., and flasks to be used many times, should be made of iron and well braced. They are then readv at all times and may be used without loss of time in repairs.

To point out the saving resulting from the use of a flask instead of bedding the pattern into the pit, the relative time recjuired for making a girder casting in the two ways may be cited. Before the flask was made to hold the pattern, it was bedded into the pit with a cope to cover it. It required a time equivalent of 14 days, with a molder and helper to complete and cast the mold. After the flask was made so the pattern was rammed in the drag and turned over, it required a time equivalent of 9 days for a molder and helper to make the same casting.

Holders' tools vary greatly with the general type of work that the molder is making. The number of tools necessary for a molder on a particular type of work may be three or four, while on intricate work manv tools may be required. There are tool manufacturers who can furnish tools of nearly any size or shape that a molder may desire. The more common forms are shown in Fig. 2. These are used for nearly all classes of work and are made in many sizes as desired. No. i is a round point finishing trowel; No. 2, a square trowel. No. 3 is a lifter for removing sand from deep and narrow parts of a mold. No. 4 is a flange and bead tool for slicking special round surfaces. Nos. 5 and 6 are two forms of double-end slickers which represent the genera; forms out of greatly varying forms of such tools. No. 5 has an oval slick at one end with the spoon slick at the other. No. 6 has the square and heart slicks. Xos. 7, 8, and 9 shoAv corner slicks of wdiich No. 7 is for round corners. No. 8 for square, and No. 9 for inside corners. Nos. 10 and 11 are pipe slicks for cylindrical surfaces. No. 10 has the square ends, while No. 11 has the safe end for a corner slick.

Fig. 2.

The success of making the mold and obtaining a

good casting is dependent mainly upon the manner of

ramming the sand to form the mold. Hard spots in the sand cause scabs, and soft spots, a swell. Uneven-

ness of ramming causes similar unevenness in the casting.

In rammino^ the drao-, the flask should be filled lo FOUNDRY PRACTICE

to a depth of from 5 to 6 inches, ramming first around the edge of the flask, then next to the pattern, and, lastly, the portion between, using the pein. On small castings, it is rarely necessary to ram the sand over the pattern. The pein or the butt of the rammer should never strike within an inch of the pattern, as it will cause a hard spot at that point.

In deep molds, the succeeding rammings should be done by filling in loose sand to the depth of about 6 inches, and ramming first with the pein then with the butt in order to give the mold the proper degree of hardness.

In ramming the drag, either the pein or the butt may be used as soon as the pattern is well covered so that the ramming is not near the pattern. It is of advantage to tramp the sand with the feet before butting, as that more quickly compresses it to a moderate hardness and facilitates the butt ramming.

In ramming the rim of a pulley, the rammer should be directed away from the pattern to prevent scabbing the rim.

The larger the pattern, the harder the sand may be rammed. When of a depth to give a great pressure on the bottom, the sand must be rammed harder to hold the pressure and prevent the cracking of the surface causing roughness sometimes called "whiskers."

On patterns of the round column type, the sand may be rammed very much harder than in other cases, if the metal is to be thick. It is very imp'^rtrmt to have these rammed evenly, as unevenness will cause defects in the casting, even though it is not of a hardness at any point which would be detrimental were the entire mold of that hardness.

The softer the sand can be left an 1 still hold the

FOUNDRY PRAC'liCE ii

casting in proper form, the less is the HabiUty of losing the casting. The sand must be hard enough to hold its shape, but after that the risk of loss in casting is increased as the hirdness is incre:;sed.

In ramming the cope where there are no bars, the sand is filled in to a depth of about 6 in. and rammed around inside the flask, then the remaining portion is rammed evenly with the pein. The butt should not be used in the cope until the entire flask is filled, then the last ramming on top is done with the butt. If the butt is used before this, it causes a hard surface so that the sand does not unite in the succeeding ramming and is lial^le to fall out when the cope is turned over. When the cope has bars, each division enclosed by the bars is rammed separately as a small cope, but all the divisions must be of an even hardness. The successive rammings are made by filling in about 6 in. at a time and ramming with the pein.

The proper manner of holding the floor rammer while ramming is to grip the rod connecting the pein and butt with one hand above the other. Never hold the rammer with one hand on top of the upper end of the rammer, as it will jar the operator and it is harder to do good ramming while in this position.

The proper venting of a mold is of as great importance as any part of the process. If at any point the venting is insuflicient to carry off the gases, the metal will blo'w and spoil the casting.

The air in the mold, when the metal is being poured, must be able to escape. This is provided for in some cases by the riser, but often the vents are depended upon for this purpose. The water in the sand is heated to steam and must escape through the sand. The continlied addition of sea coal in the facing and tiour and plumbago in the mold increases the formation of g-ases when the metal comes in contact with the face of the mold. If these can not escape into the sand, they force an opening through the molten metal which is known as blowing. To enable the gases to pass through the sand, the mold must be properly vented. Some sands are so coarse and open that they require much less venting than others which are fine and close in textiu'e.

Small and thin castings, rammed lightly, require no venting. In casting- heavy, thick plates the drag must be well vented, but the cope does not require much venting, although it is always best to A^ent it. In venting any plain casting, the size of vent wire is mainlv dependent upon the depth of sand tO' be vented. For flasks up to 12 in. in depth, Vg in. wire serves well. The exact size of wire used is unimportant, so that the venting is close enough to give free escape of the gases. The vent wire should not strike the pattern or scrape along a side, as it forms holes that the metal mav flow through and allows the metal to stop up the vent, which gives the same condition as though there were no vent. The bottom board should be put on the drag and rubbed to a bearing, then removed, and the surface creased crosswise by striking with the corner of a stick to reach the width of the drag, then the drag vented The creases form openings through which the gases may escape when the drag is turned over.

For small castings, the cope should be vented through almost to the pattern to give free escape for the gases and air. For larger castings it is often advisable to leave a layer of 2 to 3 inches next to the pattern without vents. This does not give free escape for the air ;

thus a pressure can be maintained within the mold while pouring which prevents the drawing down of the cope,

]Many molds have enclosed bodies of sand which do not have free vent connections with the top or bottom of the flask, as anchors in pulley molds, center parts in three-part flasks, or large green sand cores. The vent must be led to some convenient point where an opening is left through the cope for the gases to escape. A gutter may be cut around the surface of the body of sand about 3 or 4 inches from the pattern and connected to the vent opening. Slant vents from the gutter will give free vent to the gases.

In some molds there are pockets having metal on all sides but one. The vent must be led away through this side and this very freely. When the pocket is small, a vent rod may be laid in it, and slant vents leading to the pattern give the necessary relief. When the pocket is of large size, it is not safe to depend on the slanting vent. In these cases the gases are collected by a coke or cinder bed laid in the pocket and led to- the outside by a vent pipe or large vent rod.

In some patterns, as columns, the vent mav be led off at the parting instead of through the bottom of the drag. After the flask is rammed up and the cope removed, vents are made under the pattern from the surface of the drag at a distance of about 2 inches apart^ These are led to the outside by cutting or scratching a small gutter in the surface from the vent to the flask.

Pit molds and all floor molds must be provided with a cinder bed located about i^^ or 2 feet below the casting to provide for the escape of the gases of the lower half of the mold. The cinder bed is connected to the surface by vent pipes which give free passage for the gases. Very deep molds may have cinder beds locatcrl at different levels around the mold. The gases are led to the cinder bed by freely venting the mold so that the wire strikes the bed.

Surface molds require much better venting than those covered with a cope, as the metal gives no pressure except its weight ; thus it can not force the gases against much resistance. Small surface molds may not require venting, if the sand is rammed only enougli to prevent the metal cutting when poured. Larger molds must be provided with a cinder bed and well vented from the mold and to- the surface.

When two parts of a flask are to be lifted apart after the sand is rammed, it is necessary to make the surface at the parting so that the two bodies of sand will not knit together but separate freely when the flask is opened. In order to ensure the parting, the surface of the sand must be slicked smooth after making the surface harder than the other part of the mold and covered with a parting sand.

When the drag is turned over and the follow board is removed, the surface is gone over with the hand and sand tucked in wherever soft spots are found. The sand is then cut away to the parting of the pattern or to the surface of the part. A thin coat of sand is then riddled onto this surface and the whole slicked to a smooth face. The additional sand compacts the surface to a harder shell and should not be easily broken by the ramming oi the sand that rests on this parting. This slicking must not be carried to an extent of causing an extra hard or bricklike face, as this will cause defective castings similar to hard spots in ramming.

Parting sand is put over the surface in a thin coating. All parts of the sand mvist be covered, for any spots left bare will stick and not give a clean part. Parting sand used for this purpose may be any fine dry sharp sand, very fine cinders, or burnt core sand from the burned cores in the castings. This sand is most convenient on all plane surfaces and where the slope IS not so great that it will not stay on the entire face. In cases where the dry sand will not cover the surface well, wet sharp sand makes a good part. The fine sharp sand is dampened until the sand sticks together, then it is put onto the surface with the hand or slicked on with a tool. It often helps to dust a little dry parting sand over the wet sand after it is on the surface, as the wet sand sometimes sticks when the surfaces are lifted apart.

Parting sand and burnt core sand make molding sand coarse and weak as it loses its strength to hold a form when rammed. Too much of the parting sand will spoil the sand for the mold.

Gates are the openings through w^hich the metal enters the mold. The location of the s^ate makes a gfreat difference in the resulting casting. A mechanic can show his ability in gating properly more readilv than in any other part of the mold. Many castings are lost just because the molder is not particular enough in locating and cutting his gates.

All plain castings having about an even thickness, and that greater than the runners, are gated at the side and present little or no difficulty. As thin plates having runners heavier than the casting set sooner than the runner, they can not be gated at the side because when the runner cools the casting will be strained or warped. A good form of gate in such cases is that known as the i6 FOUNDRY PRACTICE