the peculiar motions which are best adapted for cutting the sand and removing it from the bin in a regular manner.

Another problem presents itself in the handling of core sand, after castings have been poured. Fig. 10 shows a device placed in the cleaning room of a large gray-iron foundry for the purpose of separating core rods, nails and other magnetic materials from the core sand itself. The first belt shown

in the illustration is provided with a magnetic head pulley so that the metal is dropped on the floor while the sand is delivered onto the belt which runs out of the building. The waste sand can be taken to railroad cars outside, or to a storage bin, or to trucks, or to a storage pile. The magnetic pulley does the work of several men in picking out magnetic materials, and does it cleaner and better.

Mechanical sand handling is entirely practical where the quantity of sand is great enough so that the cost of conveyor equipment, as an overhead expense is not too great. In

general it seems to be possible to put in a sand handling equipment on duplicate work where there is a minimum of 150 to 200 tons of sand to be handled daily. In some instances where the work is highly concentrated, a much smaller amount of sand than this can be handled economically.

The frequency with which molding sand may be re-used depends primarily upon the dryness of the sand before tempering, that is, if very light castings are being made and the temperature of the sand is not greatly increased, such sand can be used more frequently than that which becomes highly heated.

To my mind the problem of sand handling is simple. The less apparatus that can be used to accomplish a given result is in most cases the most satisfactory. This is due partly to the difficulty of obtaining proper mechanical attention to equipment in most foundries; but without proper mechanical attention there is no equipment that will prove satisfactory and give continuous service. In considering mechanical equipment it is extremely desirable that the question of mechanical attention be fully considered.

There is no reason why mechanical sand handling equipment cannot be produced to meet any set of conditions, providing there is sufficient sand to handle to justify the initial expense. If properly arranged, mechanical sand handling equipment should result in increased production per square foot per man on any given floor space. This has been demonstrated repeatedly. Ample proof in the form of many successful installations is available for anyone who cares to investigate.

Discussion

STAUNTON B. PECK.-In 1889 Thomas Welsh, superintendent of the Westinghouse Air Brake Co., conceived the idea of continuous molding, the molder never moving from his machine. In conjunction with the writer, a plan was worked out and the necessary conveying equipment installed the following year in the company's foundry at Wilmerding, Pa.

The molds as made were placed upon a continuously moving conveyor which carried them to the cupolas where they were poured, thence to the shake-out where the castings were picked out and taken to the mill room. The sand was riddled, tempered and elevated to an overhead conveyor running above the molding machines and delivering to them through down spouts. The returning run of the mold conveyor restored flasks, bottom boards and weights to the molders, and a separate conveyor above this brought the cores from the core room to them.

The main advantage sought was economy of labor in handling the great quantities of materials in this large foundry; but there were two other incidental features almost as valuable. The molders were relieved from the part of the ordinary molders' duties requiring least skill and were enabled to give their entire time to molding. Pouring, shaking out, tempering, etc., were performed by independent gangs, each thus attaining the highest proficiency in their special duty.

Each molder and helper put up regularly 60 flasks per hour and 600 per day; and in a foundry which had already attained a very high degree of efficiency in ordinary methods of operation and was on a piece work basis, a saving of 12c per pound was effected. This was a very large saving under the circumstances and at that time.

The other incidental advantage was the saving of room or a great reduction in the area of building needed for the required output. This can readily be appreciated as each molder required no "floor" for the accumulation of his molds before pouring. In a foundry somewhat similarly equipped some years later, I was told by the architect that the cost of the conveying system, about $80,000, was considerably less than that of the buildings required to give the additional floor area providing a conveyor system was not installed.

The Westinghouse plant was the forerunner of many other continuous or semicontinuous systems which have been installed.

The general principles are the same as described, though in some cases, where the molds are relatively large and heavy, it has been found more desirable to bring the iron to the mold

as soon as made, conveying away the large castings as fast as shaken out.

With the development of the continuous system, where the sand is used over and over again during the day, it has been found essential to give special attention to the cooling of the sand and to its preparation or conditioning.

In some cases the normal conveying of sand on belts of sufficient width to spread it thin is sufficient; in others special aerating devices are used.

The continuous use of sand is usually found to result in its becoming granular, in the formation of minute pellets, and the loss of bond. This is not peculiar to the reciprocating conveyor alone as Mr. McKinnon suggests and which others have concluded from the action of that conveyor which suggests rolling. It occurs in the same degree in systems where there are no such conveyors or even only belt conveyors. Long and close observations show it to be due to the cumulative action of chutes, screens, conveyor, riddles and molds, without time for the moisture to thoroughly and completely permeate the mass and make it perfectly homogeneous. It can be observed in ordinary molding if a relatively small amount of sand is used over and over again during the day's work. It can be wholly corrected and sand in perfect condition for molding produced by passing it through grinding pans, rolls or special devices designed for the purpose. A good example is a revivifier which is essentially a series of blades revolving at very high speed. The pellets are destroyed by impact and a perfect mixing and homogeneity obtained at the same time. It is an indispensable feature of most of the continuous systems now in successful use.

MR. G. K. HOOPER.-It always has seemed an economic wrong to laboriously handle by crude methods once a day a tonnage of sand equal to from four to ten times the tonnage of castings produced, it being apparent that by re-use during the day and by improved methods both the total quantity of sand and the labor spent upon it can be greatly reduced. The application of labor-saving methods to sand handling is, however, as in all other things, limited by the conditions surround

ing any particular product and each case must be studied upon its own merits.

The statistics given by the author as to minimum tonnage of sand to which handling methods can be profitably applied are misleading, as is the statement concerning saving of floor space. The tonnage handled has, as a matter of fact, but an indirect relation to the savings to be made, the mold being, in my judgment, the unit on which any consideration of sand handling economies should be based. This will become apparent when it is realized that the economies to be made are based on the time saved in making the mold. If the depth of the mold be constant, the length and width make but little difference, as the sand chutes can be adjusted to suit. Also a considerable difference in the amount of sand handled will make but slight difference in the installation and operation cost of a handling system. The tonnage is therefore an entirely unreliable guide.

Minimum is 5000 Molds Daily

I have in the past figured a number of proposed installations, the calculations developing that at prewar prices a production of approximately 5000 molds per day is the minimum basis for installation of such a system. At present prices of apparatus this figure would be somewhat increased, depending upon the relative increases in cost of apparatus and of labor.

Where properly applied such apparatus increases the productivity of foundry labor from 15 to 25 per cent.

In the author's statements concerning saving of floor space he has mistakenly credited the sand handling apparatus with such a saving. Any investigation of plants to which the author refers will undoubtedly reveal that some improved rapid casting method was simultaneously installed and that the saving in floor space is undoubtedly due to the latter. As a matter of fact, the installation of a sand handling system alone would require an increase in floor space since a greater number of molds could be made per day and additional floor space would be required for the increase. Floor space can be saved only by rapidly taking away the molds made.

As to the apparatus itself, it is surprising to find devices.