and being thus more completely reflected, the body appears bright and polished. It is not really the polish per se that induces the minimum of friction, but the state of the surface that causes this appearance to be presented.

As regards the frictional resistance of apparently smooth metallic surfaces, some writers have tried to explain it on the ground that metals have a surface like the nap on a felt hat. Certainly, observation seems to endorse this view, it having been noted that journals which have run in one direction for a length of time satisfactorily, get hot when run the other way, and then after a time run as cool as before -when presumably the nap, or grain, has become fairly set in the new direction. The presence of this nap in the first place may be due to the breaking up of the surface the forcible tearing apart of the fibers when machining. It is at least a coincidence that fibrous materials like wrought iron and timber evince more friction than crystalline ones such as cast iron, bronze, and babbitt.

The engineer has at different times to invoke the aid of friction and at others to overcome it. In the first case he is concerned with the action of dry surfaces, or "solid" friction; he is then desirous of arresting motion, as in brake gears, friction clutches, etc. He has to deal with the friction of lubricated surfaces, or "fluid" friction when desirous of facilitating motion, as in bearings of all kinds, such as journals, guides, glands, etc.

The laws or principles that present themselves in connection with the two kinds of friction differ considerably in some respects. To deal with solid friction first: (1) The frictional resistance is practically proportional to the

normal pressure between the two surfaces. (By "normal" is meant that the pressure is perpendicular, or square, to the surfaces; if otherwise applied, a portion of the thrust would be taken up in producing or tending to produce sliding.) (2) This resistance is, for low pressures, practically independent of the speed. (3) It depends largely on the nature of the materials in contact. (4) It is slightly greater for a state of rest than of motion. (5) It is greatest when just starting from rest, and decreases as the duration of contact elapses. This was particularly noted in the brake trials referred to; probably due to the polishing up of the surfaces, to a gradual cleansing of the contaminated surfaces, and to the particles abraded from the cast-iron brakeshoes filling up the interstices and so floating the one body over the other, exactly as graphite is intended to do nowadays when applied as a lubricant. (6) It is not greatly affected by temperature. (7) Abrading and seizing occur when the pressure becomes excessive. (8) The frictional resistance is always greater immediately after reversing the direction of the sliding motion.

The corresponding results for fluid friction are as follows: (1) When the bearing is oil-borne, the resistance is practically independent of the pressure. (This applies to bath lubrication and, in less degree, to pad and siphon lubrication also.) (2) The resistance varies directly as the speed, for low pressures. At high pressures, however, the friction is great at low velocities, thence decreases until about 100 feet per minute is reached, and thereafter increases again-about as the square root of the speed. (In connection herewith it should be remarked that actual fluid friction varies as the square of the speed; therefore, as the

friction of a thoroughly-lubricated surface is observed to vary as above mentioned, there must be involved something more than the friction of the lubricant in itself.) (3) If the bearing is flooded (the journal being oil-borne) the resistance is to a very great extent independent of the nature of the material forming the rubbing surfaces. If the lubrication falls off, however, and the surfaces come into contact, the nature of the metals asserts. itself, the condition of dry or solid friction then resulting. (4) Friction when the body is in motion is very much less than when just starting, because during the preceding state of rest, the lubricant-especially if possessing but little oiliness-will have been more or less thoroughly squeezed out, thus bringing about more nearly a metallic contact. (5) If two welloiled surfaces are brought together and then set in relative motion, the friction increases as time progresses; because the lubricating film squeezes down under pressure. Also, the effect that Also, the effect that brings about the reduction in the case of dry friction is not here present, because the surfaces not being in contact, there is no cause at work to alter their nature. (6) The resistance varies with the changing of the temperature,

with 600 pounds per square inch before failing. The limit of pressure depends on the viscosity or rather the oiliness of the lubricant. (8) The same effect is observed as in dry friction. It is thought that the metal

-however highly polished to the eyehas a grain in one direction, and that when running against the grain (like rubbing the nap on a hat the wrong way) the friction is necessarily higher. After a few hours of the new direction, the friction gets back to the original amount-having in that interval presumably set the nap the other way. The term "coefficient of friction" is an ever-recurring one in connection with any discussion of the present subject. If we place a block of wood weighing 50 pounds on a wooden table and attach a spring balance to it, and exert a steady pull on the same, it will register, say, 20 pounds (Fig. 3). Then the ratio of the pull to the dead 20 weight is or .4. Also, if we tilt the 50 table up steadily, the block will, after a certain elevation is reached, move down it (Fig. 4). The angle of tilt will be

This ratio, which by the way, is a measure of the angle the function known as the tangent, in fact—is the same as found above. The angle in question is called the angle of repose for the particular materials employed; it is also spoken of as the "friction angle." If we were to substitute metallic surfaces, smooth and oiled, the ratio of the pull to the dead weight, would be about .1. The angle of repose will be found to be about 6 degrees and the

tangent of this angle is. If we were to hang a weight over a pulley, as in Fig. 5, the weight required to move the block would be 20 pounds, ignoring the friction of the pulley axle itself.

Another point we should notice, and one that endorses Law 4, is that after the weight is once started, the weight recorded on the spring balance wouldif the same rate of motion were maintained-drop below the 20-pound mark, and this, too, independent of any inertia effects. Similarly if, while the weight was falling, we could arrange to continually take off minute amounts every instant, the velocity would be maintained constant; if we kept the same weight on, the velocity would be continuously

accelerated. Here again, we would remark, this result would ensue, over and above the acceleration due to gravity.

In regard to fluid friction, it may occur to the reader that the friction of a body in water is less than in oil, and such is actually the case. The ordinary textbook experiment is to suspend a hollow cylinder inside a vessel filled with water; if the supporting wire be twisted through a certain angle and then released, the cylinder will swing around until the twist is unwound and the wire twisted up a certain amount in the other direction. It will thus vibrate, twisting and untwisting, until it comes to rest. Obviously the greater the friction all around the sides of the cylinder, the sooner will the latter come to rest. If, then, for the water we substituted oil or, worse still, molasses or tar, the sooner should we expect the motion to cease; and this in fact is borne out by experiment. Since then, water opposes less frictional resistance to the motion of bodies in contact with it, why not employ it. in bearings? The answer which is readily deduced is that water has practically no power of maintaining a film under pressure; it would quickly squeeze out and allow the surfaces to come into contact. Thus the desiderata in a lubricant are: an ability to maintain a film under high pressure and temperature, and at the same time cause but little internal friction and, we may add, also possess good heatconducting qualities.

In the next issue we shall say something about frictional work and heating; about loss of efficiency in machines; and about the general statement-an untrue one-that friction always decreases with the rubbing speed.

(To be Concluded.)

BEFOR

EFORE entering upon the subject of the capacity of steam boilers, it is well to consider a few points relative to steam making, efficiency, and the requirements of a perfect boiler.

The prime requisite of a boiler is to furnish a quantity of energy in the form of measurable heat and molecular activity. The technical term for this combination is "latent heat."

Suppose that we should take a pound of ice at the temperature of absolute zero, equivalent to 460° below the zero of Fahrenheit. Now add heat to this ice until it has reached a temperature of 492° absolute or 32° F. From this point the addition of heat makes no sensible change in the temperature of the ice, until enough. heat has been added to have made the ice 283° hotter, if this were possible. As it is not possible, the heat is absorbed by the ice in such a manner that, although melted, the temperature of the resultant water remains at 32° F. The moment that 283° of heat have been added, the water commences to increase in sensible heat for every degree above that, until it has reached a temperature of 672° absolute, or or 212° F.

We have now reached another critical point and can add heat to the water without increasing its measurable temperature until enough heat has been added to have raised the thermometer 966° or to 1,178° F. All the water has now become steam, although still at the sensible heat of 212° F.

About four-fifths of the heat added to the pound of ice has now disappeared and is not sensible to any of our instruments for measuring heat.

From this point if heat be added it increases the temperature in a ratio proportional to the pressure and volume.

It is not known if there is any other critical point, until the component parts of the steam are separated into their original gases; that is, oxygen and hydrogen.

The heat, which is stored in the water and steam, and which is insensible to our our thermometer, is called "latent heat," and on this heat depends the ability of the steam to do work.

Our position at present, then, is this: A given quantity of water at a given temperature has been converted into steam at a given temperature by having a certain amount of heat applied to it, and that heat is now ready to appear, by a reverse of the process, in the form of power at the steam engine.

It is impossible to utilize all the heat given out by the coal for the reason that a portion of the products of combustion escape through the chimney or smoke pipes, the temperature of the hot gases at that point ranging from 600° to 1,000° F.; although the last named figure usually indicates a poorly designed uptake and stack. Another portion of the gases is lost in radiation from the outside of the boiler before reaching the water, and from the pipes after the steam has left the boiler.

The portion which penetrates the metal heating surface of the boiler, and which is absorbed by the contiguous water, is the only part of the heat that is useful to us. The loss of heat at the chimney will generally run from 20 to 40 per cent.

One pound of pure carbon when burned will yield an average of 14,500

heat units, each equal to 778 ft. lbs. of work. If all this heat could be utilized in one hour it would exert about 5.697 horsepower. As it is, under present conditions of utilization of the heat from coal, we can only realize in one hour from one-quarter to one-half of a horsepower. The number of heat units equivalent to this would evaporate about 15 lb. of water from 212° F. at atmospheric pressure.

A good marine or land boiler of today will evaporate about 9 lb., therefore but about 60 per cent. of the heat is used.

The rule for finding the efficiency of a boiler is as follows (see "Foley's Mechanical Engineers' Reference Book"):

"Given the heat value of the fuel in thermal units per pound, and the number of pounds of water actually evaporated per pound of fuel burned; also the temperature of the feedwater; to find the efficiency of performance the initial steam pressure being known."

HEATING VALUE OF AMERICAN COAL,
BABCOCK & WILCOX

of water from and at 212° per pound of combustible is meant that the evaporation of water is considered to have taken place at mean atmospheric pressure, and at the temperature due to that pressure, the feedwater also being assumed to have been supplied at the same temperature; this is equivalent to 965.7 B. T. U. per pound of water. An example of the use of the above formula follows:

Find the efficiency of a boiler which evaporates one pound of water per pound of coal at 85 lb. pressure (100 lb. absolute). The feed supplied at 120° F. and fuel average American bituminous coal.

Total heat from 0° F. at 100 lb. absolute 1213.8
Temperature of feed

Supplied for each pound of steam

120.

1093.8

See "Steam Table" in textbooks under heading "Total Heat From Ordinary Zero F. British Units."

Therefore, 1093.8 x 9 = 9844.2 units used out of the 14,750 in each pound of coal. Then to get the efficiency divide 9844.2 by 14,750, which gives.667.

Our boiler has 66.7 per cent. efficiency, which is pretty near the limit for a good boiler. The requirements of a perfect boiler, partly from Mr. Babcock's lectures, are as follows:

First.-Best material, simple in construction, perfect in workmanship and durable in use.

Second. Steam and water capacity sufficient to prevent any fluctuation in pressure or water level.

Third.-A large water surface for the disengagement of the steam from the water to prevent foaming.

Fourth.-Sufficient steam room.