5. From the saw teeth having insufficient set, and thus causing the saw to heat.
The methods of discovering the errors of tension in a saw, and the process of hammering to correct them, have already been explained with reference to the use of the hammer on pages from [68] to [70] of volume 2 of this work.
Before hanging a saw on a mandrel, it is necessary to know that the mandrel itself runs true in its bearings or boxes. In a new machine this may be assumed to be the case, but it is better to know that it is so, because if the mandrel does not run true several very improper conditions are set up. First, the saw will run out of true circumferentially, and therefore out of balance, and the high side of the saw will be called upon to do more cutting duty than the low side. Second, the centrifugal force will be greatest on the high side, and the saw will be stiffer, thus setting up an unequal degree of tension. Third, the saw will run out of true sideways, cutting a wider kerf than it should, thus wasting timber while requiring more power to drive.
The collar on the saw arbor should be slightly hollow, so that the saw will be gripped around the outer edge of the collar, and the arbor or mandrel should be level so that the saw will stand plumb. The boxes or bearings of the arbor should be an easy working fit to the journals, and there should be little, or what is better, no end play of the arbor in its bearings.
If a saw arbor becomes heated enough to impair the tension of the saw, it has been hot enough to impair its own truth, and should be examined and trued if necessary.
The most important point in this respect is that the face of the collar against which the saw is clamped should run true, bearing in mind that if it is one hundredth of an inch out of true in a diameter of, say 3 inches, it becomes twenty hundredths or one-fifth of an inch at the circumference of a saw that is 60 inches in diameter.
In cases of necessity, a saw that wabbles from the collar face of the mandrel running out of true, may be set true by means of the insertion of pieces of paper placed between the saw and the face of the collar.
The first thing to do in testing the saw is to take up the end motion of the saw arbor, or if this cannot be done, then a pointed piece of iron or wood should be pressed on the end of the mandrel so as to keep it from moving endways while the saw is being tested.
The saw should be revolved slowly, and a piece of chalk held in the cleft of a piece of wood should be slowly advanced until it meets some part of the face of the saw just below the bottom of the saw teeth.
As soon as the chalk has touched and the saw has made one or two revolutions the chalk should be moved a trifle farther on from the teeth, and another mark made, and then moved on again, and so on, care being taken to notice how much space there is between the high and low sides of the saw. It will be found, however, that the shorter the chalk marks are the more the saw is out of true.