[Ill.u.s.tration: FIG. 31.--Static bending test on small beam. Note the use of the deflectometer with indicator and dial for measuring the deflection; also roller bearings between beam and supports.]

_Preparing the material_: The specimens may be of any convenient size, though beams 2" X 2" X 30" tested over a 28-inch span, are considered best. The beams are surfaced on all four sides, care being taken that they are not damaged by the rollers of the surfacing machine. Material for these tests is sometimes cut from large beams after failure. The specimens are carefully weighed in grams, and all dimensions measured to the nearest 0.01 inch. If to be tested in a green or fresh condition the specimens should be kept in a damp box or covered with moist sawdust until needed. No defects should be allowed in these specimens.

_Marking and sketching_: Sketches are made of each end of the specimen to show the character of the growth, and after testing, the manner of failure is shown for all four sides. In obtaining data regarding the rate of growth and the proportion of late wood the same procedure is followed as with large beams.

_Adjusting specimen in machine_: The beam should be correctly centred in the machine and each end should have a plate with roller bearings between it and the support. Centre loading is used. Between the movable head of the machine and the specimen is placed a bearing block of maple or other hard wood, the lower surface of which is curved in a direction along the beam, the curvature of which should be slightly less than that of the beam at rupture, in order to prevent the edges from crushing into the fibres of the test piece.

_Measuring the deflection_: The method of measuring deflection of large beams can be used for small sizes, but because of the shortness of the span and consequent slight deformation in the latter, it is hardly accurate enough for good work. The special deflectometer shown in Fig. 31 allows closer reading, as it magnifies the deflection ten times. It rests on two small nails driven in the beam on the neutral plane and vertically above the supports. The fine wire on the wheel at the base of the indicator is attached to another small nail driven in the beam on the neutral plane midway between the end nails. All three nails should be in place before the beam is put into the machine. The indicator is adjustable by means of a thumb-screw at the base and is set at zero before the load is applied.

Deflections are read to the nearest 0.001 inch. For rate of application of load see SPEED OF TESTING MACHINE, above. The speed should be uniform from start to finish without stopping.

Readings must be made "on the fly."

_Log of the test_: The log sheets used for small beams (see Fig.

32) are the same as for large sizes and the procedure is practically identical. The stress-strain diagram is continued to or beyond the maximum load, and in a portion of the tests should be continued to six-inch deflection or until the specimen fails to support a load of 200 pounds. Deflection readings for equal increments of load are taken until well beyond the elastic limit, after which the scale beam is kept balanced and the load read for each 0.1 inch deflection. The load and deflection at first failure, the maximum load, and any points of sudden change should be shown on the diagram, even though they do not occur at one of the regular points. A brief description of the failure and the nature of any defects is entered on the log sheet.

[Ill.u.s.tration: FIG. 32.--Sample log sheet, giving full details of a transverse bending test on a small pine beam.]

_Calculating the results_: The formulae used in calculating the results of tests on small rectangular simple beams are as follows:

0.75 P (1) J = -------- b h

1.5 P_{1} l (2) r = ------------- b h^{2}

1.5 P l (3) R = --------- b h^{2}

P_{1} l^{3} (4) E = ------------- 4 D b h^{3}

P_{1} D (5) S = --------- 2 V

The same legend is used as in BENDING LARGE BEAMS. The weight of the beam itself is disregarded.

ENDWISE COMPRESSION

_Apparatus_: An ordinary static testing machine and a compressometer are required. (See Fig. 33.)

[Ill.u.s.tration: FIG. 33.--Endwise compression test, showing method of measuring the deformation by means of a compressometer.]

_Preparing the material_: Two cla.s.ses of specimens are commonly used, namely, (1) posts 24 inches in length, and (2) small clear blocks approximately 2" X 2" X 8". The specimens are surfaced on all four sides and both ends squared smoothly and evenly. They are carefully weighed, measured, rate of growth and proportion of late wood determined, as in bending tests. After the test a moisture section is cut and weighed. Ordinarily these specimens should be free from defects.

_Sketching_: Sketches are made of each end of the specimens to show the character of the growth. After testing, the manner of failure is shown for all four sides, and the various parts of the failure are numbered in the order of their occurrence.

_Adjusting specimen in machine_: The compressometer collars are adjusted, the distance between them being 20 inches for the posts and 6 inches for the blocks. If the two ends of the blocks are not exactly parallel a ball-and-socket block can be placed between the upper end of the specimen and the movable head of the machine to overcome the irregularity. If the blocks are true they can simply be stood on end upon the platform and the movable head allowed to press directly upon the upper end.

_Measuring the deformation_: The deformation is measured by a compressometer. (See Fig. 33.) The latter registers to 0.001 inch. In the case of posts the compression between the collars is communicated to the four points on the arms by means of bra.s.s rods; with short blocks, as in Fig. 33, the points of the arms are in direct contact with the collars. The operator lowers the fulcrum of the apparatus by moving the micrometer screws at such a rate that the set-screw in the rear end of the upper lever is kept barely touching the fixed arm below it, being guided by a bell operated by electric contact.

_Log of the test_: The load is applied continuously at a uniform rate of speed. (See SPEED OF TESTING MACHINE, above.) Readings are taken from the scale of the compressometer at regular increments of either load or compression. The stress-strain diagram is continued to at least one deformation point beyond the maximum load, and in event of sudden failure, the direction of the curve beyond the maximum point is indicated. A brief description of the failure is entered on the log sheet. (See Fig. 34.)

[Ill.u.s.tration: FIG. 34.--Sample log sheet of an endwise compression test on a short pine column.]

In short specimens the failure usually occurs in one or several planes diagonal to the axis of the specimen. If the ends are more moist than the middle a crushing may occur on the extreme ends in a horizontal plane. Such a test is not valid and should always be culled. If the grain is diagonal or the stress is unevenly applied a diagonal shear may occur from top to bottom of the test specimen. Such tests are also invalid and should be culled. When the plane (or several planes) of failure occurs through the body of the specimen the test is valid. It may sometimes be advantageous to allow the extreme ends to dry slightly before testing in order to bring the planes of failure within the body. This is a perfectly legitimate procedure provided no drying is allowed from the sides of the specimen, and the moisture disk is cut from the region of failure.

_Calculating the results:_ The formulae used in calculating the results of tests on endwise compression are as follows:

P (1) C = ----- A

P_{1} (2) c = ------- A

P_{1} l (3) E = --------- A D

P D (4) S = ----- 2 V

C = crushing strength, pounds per square inch.

c = fibre strength at elastic limit, pounds per square inch.

A = area of cross section, square inches.

l = distance between centres of collars, inches.

D = total shortening at elastic limit, inches.

V = volume of specimen, cubic inches.

Remainder of legend as in BENDING LARGE BEAMS, above.

COMPRESSION ACROSS THE GRAIN

_Apparatus_: An ordinary static testing machine, a bearing plate, and a deflectometer are required. (See Fig. 35.)

[Ill.u.s.tration: FIG. 35.--Compression across the grain. Note method of measuring the deformation by means of a deflectomoter.]

_Preparing the material_: Two cla.s.ses of specimens are used, namely, (1) sections of commercial sizes of ties, beams, and other timbers, and (2) small, clear specimens with the length several times the width. Sometimes small cubes are tested, but the results are hardly applicable to conditions in practice. In (2) the sides are surfaced and the ends squared. The specimens are then carefully measured and weighed, defects noted, rate of growth and proportion of late wood determined, as in bending tests. (See BENDING LARGE BEAMS, above.) After the test a moisture section is cut and weighed.

_Sketching_: Sketches are made as in endwise compression tests.

(See ENDWISE COMPRESSION, above.)

_Adjusting specimen in machine_: The specimen is laid horizontally upon the platform of the machine and a steel bearing plate placed on its upper surface immediately beneath the centre of the movable head. For the larger specimens this plate is six inches wide; for the smaller sizes, two inches wide. The plate in all cases projects over the edges of the test piece, and in no case should the length of the latter be less than four times the width of the plate.

_Measuring the deformation_: The compression is measured by means of a deflectometer (see Fig. 35), which, after the first increment of load is applied, is adjusted (by means of a small set screw) to read zero. The actual downward motion of the movable head (corresponding to the compression of the specimen) is multiplied ten times on the scale from which the readings are made.

_Log of the test_: The load is applied continuously and at uniform speed (see SPEED OF TESTING MACHINE, above), until well beyond the elastic limit. The compression readings are taken at regular load increments and entered on the cross-section paper in the usual way. Usually there is no real maximum load in this case, as the strength continually increases as the fibres are crushed more compactly together.

_Calculating the results_: Ordinarily only the fibre stress at the elastic limit (c) is computed. It is equal to the load at elastic limit (P_{1}) divided by the area under the plate (B).

{ P_{1} } { c = ------- } { B }

SHEAR ALONG THE GRAIN