In the installation, maintenance and repair of shafting, as in all other things, there is a right and a wrong way; and though the wrong way ranges in its defects from matters causing trivial inconvenience to absolute danger, the right too often—owing to lack of knowledge or discernment—finds but scant appreciation.

Where, as is often the case, the end of a shaft is journaled to admit of the use of an odd, small-bore pillow block or wall-box hanger, the journaled part should equal in length twice the length of the hanger bearing plus the length of the collar. The hanger can thus readily be slid out of the wall box, and the necessity of uncoupling this shaft length and removing it before access to the bearing for purposes of cleaning or repair is done away with.

A plank or board A (Fig. 1), about ¼ to ½ inch longer than the distance from the bottom of the shaft to the floor, can be used to good advantage at such times to free the hanger of the shaft's weight, and to prevent the shaft's springing from its own weight and the pulleys it may be carrying.

Should it become necessary to place a pulley with half the hub on and half off the journaled part, this can readily be done by the use of a split bushing, as shown in sectional view of Fig. 1.

Very often a small-sized bearing is used and the shaft journaled off to act as a collar. Of this procedure it can only be said that if done with the idea of making a "good job" it signally fails of its object; if of necessity (a collar being insufficient), then the shaft is heavily overloaded and serious trouble will result, because of it.

It is advisable to center punch, or otherwise mark, the ends of both shafts held by a compression coupling close up against the coupling, and both edges of the coupling hub should have a punch mark just opposite and close to the shaft punch marks. These marks will serve at all times to show at a moment's glance any end or circumferential slippage of the shafts within the coupling. The same method can be resorted to for proof of pulley slippage.

When a new line of shafting is put up, the foot position of each hanger should be clearly marked out on their respective timbers after the shaft has been brought into alinement. Hangers can thus be easily put back into their proper place should timber shrinkage or heavy strains cause them to shift out of line. This idea can be applied to good advantage on old lines also, but before marking out the hanger positions the shaft should be tried and brought into perfect alinement.

Hangers that do not allow of any vertical adjustment should not be used in old buildings that are liable to settle. Shafting so run pretty nearly always gets out and keeps out of level.

In flanged bolt couplings (Fig. 1) no part of the bolt should project beyond the flanges. And where a belt runs in close proximity to such a coupling, split wood collars should be used to cover in the exposed coupling flanges, bolt heads and nuts. Countershafts have been torn out of place times innumerable by belts getting caught and winding up on the main line.

Whenever possible a space of 8 to 10 inches should be left between the end of a shaft line and the wall. A solid pulley or a new coupling can thus readily be put on by simply uncoupling and pushing the two shaft lengths apart without taking either down. Ten inches does not represent the full scope of pulleys admissible, for so long as the pulley hub does not exceed a 10-inch length the pulley face (the more readily in proportion to the larger pulley diameter) can be edged in between the shafts.

Fig. 2 is an instance of bad judgment in locating the bearings. In one case this bearing overheated; the remedy is either to re-babbitt the old box or replace it with a new one.

Both pulleys were solid and the keys—headless ones—had been driven home to stay. The rims of both pulleys almost touched the wall, and the circumferential position on the shaft of both these pulleys was such as to preclude the possibility (owing to an arm of a being in a direct line with key B¹ and arm of b with key a¹) of using anything but a side offset key starting drift.

An effort was made to loosen b (which was farthest from the wall) by sledge-driving it toward the wall, hoping that the pulley might move off the key. The key, as was afterward found out, not having been oiled when originally driven home had rusted in place badly; though the pulley was moved by sledging, the key, secure in the pulley hub, remained there.

Ultimately one of us had to get into pulley b, and, removing cap c, hold the improvised side offset, long, starting drift D in place against B¹ at b² while the other swung the hand sledge at a. The entering end of the key, not having been file chamfered off, as it should have been (see E), our starting drift burred it up; so, after having started it, we had the pleasure of getting into b to file the key end b² into shape so as to admit of getting it out.

The solid pulley b has since been replaced with a split pulley.

By the arrangement, as shown in Fig. 3, of the rim-friction clutch on the driven main shaft B and the driving pulley on the engine-connected driving main shaft A, no matter whether B shaft is in use or not—i.e., whether the clutch be in or out of engagement—so long as A shaft is in motion the belt C is working.

Main line belts come high, and the more they are used the sooner will they wear out. By changing the clutch from shaft B to A and the pulley D from A to B, belt C will be at rest whenever B is not in use. Where, however, these shafts are each in a separate room or on a different floor (the belt running through the wall or floor and ceiling, as the case may be) the clutch, despite belt wear, should be placed directly on the driven shaft (as B), so as to provide a ready means for shutting off the power in cases of emergency.

Figs. 4, 5 and 6 represent a dangerous mode, much in vogue, of driving an overhead floor. An extremely slack belt connects the driving shaft A and the driven shaft B; when it is desired to impart motion to the driven shaft the belt tightener C is let down and belt contact is thus secured.

This tightener system is called dangerous advisedly, for few are the shops employing it but that some employee has good cause to remember it. Unlike a clutch—where control of the power is positive, instantaneous and simple—the tightener cannot be handled, as in emergency cases it has to be.

In any but straight up and down drives with the driven pulley equal to or larger (diametrically) than the driver, unless the belt have special leading idlers there is more or less of a constant belt contact with its resultant liability to start the driven shaft up unexpectedly. When the tightener is completely off, the belt, owing to heat, weight or belt fault, may at any time continue to cling and transmit power for a short space, despite this fact.

These tighteners are usually pretty heavy—in fact, much heavier than the unfamiliar imagines when on the spur of emergency he grapples them, and trouble results.

Tightener (in Fig. 5) A is held in place by two threaded rods B—as shown by slot a in A¹—and regulated and tightened by ring-nuts C working along the threaded portion of B. C (of Fig. 4) is also a poor arrangement. Fig. 6 is the best of them all.

Apropos of clutches, great care must be exercised in tightening them up while the shafting is in motion, for if the least bit overdone the clutch may start up or, on being locked for trial (according to the clutches' structure), continue running without possibility of release until the main source of power be cut off. Nothing can exceed the danger of a clutch on a sprung shaft.

Heavily loaded shafting runs to much better advantage when center driven than when end driven, and what often constitutes an overload for an end drive is but a full load for a center drive. To illustrate, here is one case of many: The main shaft—end driven—was so overloaded that it could be alined and leveled one week and be found out one way or the other, frequently both ways, the next week. Being tired of the ceaseless tinkering that the condition under which that shaft was working necessitated, the proprietors were given the ultimatum: A heavier line of shafting which would be sure to work, or a try of the center drive which, owing to the extreme severity of this case, might or might not work.

A center drive, being the cheapest, was decided upon. Pulley A, Fig. 7, which happened to be a solid, set-screw and key-held pulley, was removed from the end of the shaft. The split, tight-clamping-fit pulley B, Fig. 8, was put in the middle of the shaft length; the gas engine was shifted to accommodate the new drive, and hanger C¹ was put up as a reinforcement to hanger C and as a preventive of shaft springing. After these changes the shaft gave no trouble, so that, as had been hoped, the torsional strain that had formerly all been at point 1 must evidently have been divided up between points 2 and 3.

When a main shaft is belted to the engine and to a countershaft, as shown in Fig. 9, the pulley A¹ gets all the load of main and countershafts. In the arrangement shown in Fig. 10 point 1 gets A's load and 2 gets B's load and is the better arrangement.

Where a machine is situated close to one of the columns or timber uprights of the building it is very customary to carry the belt shifter device upon the column, as in Fig. 11. The sudden stoppage of a machine seldom does any damage, whereas an unexpected starting may cause irreparable damage and often even endanger the limb and life of the machine operative.

To avoid the possibility of some passing person brushing up against the shifting lever and thus starting the machine, the tight and loose pulleys of the countershaft should be so placed that when A is exposed—that is, away from the column—its accidental shifting shall stop the machine. Fig 12 makes this point clear.

This arrangement is often used to save a collar (at A). The oil runs out between the loose pulley and the bearing, especially if the latter be a split bearing; the loose pulley, instead of being totally free when the belt is on the tight pulley, acts more or less, in proportion to the end play of the shaft, as a buffer between the tight pulley and the bearing; finally, the tight pulley is deprived of the support (which, when under load, it can use to good advantage) a nearer proximity to the hanger would give it.

The shafts of light-working counters should not be needlessly marred with spotting or flats for collar set-screws, nor should cup or pointed set-screws (which mar a shaft) be used. If the collar be sharply tapped with a hammer, diametrically opposite the set-screw, while it is being tightened up, all slack is taken out of the collar; and the hold is such that, without resource to the same expedient when loosening the collar, a screwdriver will scarcely avail against a slotted set-screw.

When required to sink the head of a bolt into a timber to admit of the timbers lying snug in or against some spot, if allowable, the bolt's future turning can be guarded against by cutting the hole square to fit the bolt head. But where a washer must be used, the only positive and practical way to prevent the bolt from turning is to drive a nail (as shown) into A (Fig. 13) far enough for the nail head to flush B; now bend the head down behind the bolt toward c. It is evident that if the bolt tries to turn in the direction of 3 the nail end (wood held) will prevent it; if toward 4, the nail head will be forced against the wood and catch hold of the bolt head.

Large belts of engines, dynamos, motors, etc., when in need of taking-up are usually attended to when the plant is shut down; that is, nights, Sundays or legal holidays. At such times power is not to be had; and if the spliced part of the belt, which must be opened, shortened, scraped, re-cemented and hammered, happens to be resting against the face of one of the pulleys, is up between some beams or down in a pit, the chances of the job, if done at all, being any good are very slim.

The spliced part of a large belt should be clearly marked in some permanent and easily recognizable way (a rivet, or where the belt is rivet-held at all its joints some odd arrangement of rivets is as good a way as any). This marking will minimize the possibility of mistake and enable the engineer to place the belt splice in the position most favorable for the belt-maker's taking-up.

In wire-lacing a belt, very often, despite all efforts and care, the edges of the belt (A, B) get out of line, as shown in Fig. 14, and make the best of jobs look poor. By securing the belt in proper position by two small pieces of wire passed through and fastened at 1, 2, 3 and 4, Fig. 15, the lacing can be more conveniently accomplished and the edge projection is avoided. When the lacing has progressed far enough to necessitate the removal of wires c d, the lacing already in place will keep the belt in its original position.

A wire lacing under certain conditions will run a certain length of time to a day. On expensive machinery whose time really is money it pays to renew the lacing at regular intervals so as to avoid the loss of time occasioned by a sudden giving out of the lace.

Never throw a belt on to a rim-friction or other kind of clutch while the shaft is in full motion. Belts, when being thrown on, have a knack, peculiarly their own, of jumping off on the other side of the pulley. And should a belt jump over and off on the wrong side and get caught in the clutch mechanism, as the saying goes, "there will be something doing" and the show usually comes high. It pays to slow down.

A mule belt (transmitting in the neighborhood of or considerably over 25 horse-power) that runs amuck through the breaking down of the mule can make enough trouble in a short time to keep the most able repairing for a long while.

No matter what the pulley shafts holding arrangement and adjusting contrivance may be, all of the strain due to belt weight, tension, and the power transmitted falls mainly at points A, A¹, Fig. 16; and it is here that, sooner or later, a pin, set-screw or bolt gives way and the belt either gets badly torn up, rips something out of place, or a fold of it sweeping to the floor slams things around generally until the power is shut off.

The remedy is obvious: Reinforce A, A' by securing B, B' to the supporting shaft c at c¹, c². The yoke x is a reliable and practical means to this end. Straps a held by the nuts b hold the yoke securely on the supporting shaft c, while the pulley-shaft ends B, B' are held in the U of the yoke at w' at any desired distance from c by means of the adjustment provided by the nuts b.

The end of a hanger bearing was badly worn (Fig. 17). The cap could be lifted out by removing bridge A, but the shaft interfered with the lifting of the bottom out, owing to its being held in the hanger slides. It had to be removed and we were called upon to put it into shape by re-babbitting.

Being a newspaper plant, money was no object; the time limit, however, was three hours, or hands off. Opening the 30-inch engine belt and removing the interfering shaft length was out of the question in so short a time. So the job was done as follows: The shaft was braced against down sag and engine pull along the line B C by a piece of timber at A, and against pull on B D by timber arrangement X; timber y's points y¹ and y² resting against the uprights at 1 and 2, timber z wedged in between y at y³ and the shaft at 4, thus acting as the stay along line B D. The nuts and washers a, a were removed; the bolts driven back out of the bracket; the end of a rope was thrown over the shaft at b, passed through the pulley and tied to the bracket and hanger which, as one piece, were then slid endways off the shaft and lowered to the floor. The bearing was cleaned, re-babbitted and scraped, everything put back, stays removed and the shaft running on time with a half-hour to the good.

When desirable to keep a shaft from turning while chipping and filing flats, spotting in set screws or moving pulleys on it, it can be done by inserting a narrow strip of cardboard, soft wood or several thicknesses of paper between the bearing cap and the top of the shaft and then tightening the cap down.

The packing, 1/16 to 3/16 inch thick and about as long as the bearing, must be narrow; otherwise, as may be deduced from Fig. 18 (which shows the right way), by the use of a wide strip in the cap the shaft is turned into a wedge, endangering the safety of the cap when forced down. At point 3 packing does no harm, but at 1 and 2 there is just enough space to allow the shaft diameter to fit exactly, with no room to spare, into the cap bore diameter.

As a very little clamping will do a good deal of holding the clamping need not be overdone. A shaft can also be held from turning, or turned as may be desired, by holding it with a screw (monkey) wrench at any flat or keyway, as shown in sectional view, Fig. 19.

When a shaft breaks it is either owing to torsional strain caused by overload, springing through lack of hanger support at the proper interval of shaft length, the strain of imperfect alinement or level, or a flaw.

An immediate temporary repair may be effected by taking some split pulley that can best be spared from another part of the shaft and clamping it over the broken part of the shaft, thus converting it, as it were, into a compression coupling. The longer the pulley hub the better the hold; spotting the set-screws—that is, chipping out about 1/8-inch holes for their accommodation into the shaft—is also a great help.

If when the shaft breaks it has not been sprung by the sudden dropping of itself and the pulleys that were on it, a permanent repair can be effected, after correcting the cause of the break, by the use of a regular key-less compression coupling.

If it has been sprung, a new length comes cheapest in the wind-up; and if overload was the original cause of the trouble, only a heavier shaft or a considerable lightening of the load will prevent a repetition.

In Fig. 20 A shows how to drive to make belt weight count in securing extra contact. In B this weight causes a loss of contact. Bearing in mind that B is not only a loss from the normal contact but also a loss of the extra contact that A gives, it will readily be seen how important a power-saving factor the right sort of a drive is—especially on high-speed small-pulley machines, such as dynamos, motors, fans, blowers, etc.

A good many electrical concerns mount some of their styles of dynamos and motors (especially the light duty, small size) upon two V-shaped rails, Fig. 21 (the bottom of the motor or dynamo base being V-grooved for the purpose). The machine's weight and the screws A are counted on to keep it in place. If the machine be properly mounted on these rails, as regards screws A in relation to its drive, the screws reinforce the machine's weight in holding it down and also permit a surer adjustment through this steady holding of the machine.

Fig. 22 shows the machine properly mounted. The belt tension and pull tend to draw B corner of the machine toward the shaft C; and screw B¹ is there to resist this pull. Owing to this resistance and the pull along line D, E tends to lift and slew around in E¹ direction; screw E² is, however, in a position to overcome both these tendencies. If the screws are both in front, there is nothing but the machine's weight to keep the back of it from tilting up. The absurdity of placing the screws at F and G, though even this is thoughtlessly done, needs no demonstration.

When putting a new belt on a motor or dynamo, both the driver and the driven are often needlessly strained by the use of belt-clamps, in the attempt to take as much stretch out of the belt as possible. On being loosely endlessed it soon requires taking up; and if only laced, when the time for endlessing comes the belt is botched by the splicing in of the piece which, owing to the insufficiency of the original belt length, must now be added to supply enough belt to go around, plus the splice.

The proper mode of procedure is: Place the motor on its rails or slides 5 inches away from its nearest possible approach to the driven shaft or machine and wire-lace it (wire-lacing is a very close second to an endless belt). Let it run for a few days, moving the motor back from the driven shaft as the belt stretches. When all reasonable stretch is out, move the motor back as close to the driven shaft as possible.

The 5 inches forward motion will give 10 inches of belting, which will be amply sufficient for a good splice; and, further, the machine will be in position to allow of tightening the belt up, by simply forcing the motor back, for probably the belt's lifetime.

---

II