P. A. Kozmin.

STUDY OF THE STRENGTH

OF YELLOW PINE

(PINUS PALUSTRIS).

Tomsk.

Steam Typo-Lithography of P. I. Makushin, Blagoveshchensky Lane, own building.

1905.

General view of yellow pine (Pinus palustris).

Study of the Strength of Yellow Pine (Pinus palustris).

Experiments conducted by Engineer-Technologist Kozmin in the Washington laboratory of the Forest Department of the United States of North America.

I.

Industry in the United States of North America developed almost entirely in isolation from European industry, independently working out original technical methods of production in the first, and partly at the beginning of the second, half of the past century.

The distinctive genius of American invention, of the so-called “self-made engineers” (that is, self-taught engineers), of whom Thomas Edison is a typical representative, had an enormous influence on the development of the technology of the manufacturing industry of the United States.

Cut off by the ocean from Europe (until the middle of the past century), this country, richest in natural resources, followed a wholly distinctive path in technical improvement. And if one takes into account the circumstance that Western Europe was already then resounding with great names of technical science, then the exceptionally contemptuous attitude of European engineers toward their overseas colleagues will be understandable—an attitude which, however, did not last long—only until 1876. In 1876, at the World’s Fair in Philadelphia, European engineers made the “discovery of America”—there was so much that was new and original for them in the field of American machine building—and from that time the attitude of learned European technologists toward the American “self-taught engineers” changed. At present, after the exhibition, the copying of more or less successful American designs began, continuing up to the present moment, when, indeed, the European designer has something to learn from the Americans. But that same exhibition had, to some extent, a fatal influence on the development of steam technology in the United States, owing to the fact that the enthusiastic reviews by European authorities in technology about the American self-taught engineers gave grounds for every kind of self-

may be attributed to its characteristic disregard for the scientific foundations of engineering. True, the technical genius of American constructors has accomplished much in the field of machine building, but the path of guesswork has often proved far from inexpensive precisely in those fields where knowledge of the properties of one or another construction material is required—knowledge that is attained only through the rational organization of laboratory tests. And since the beginning of the development of higher technical education should be assigned to the 1880s of the last century*), the testing of materials—scientifically grounded, of course—also belongs to that same period.

Wood has fared worst of all in this respect. In order to judge how deeply technical “quackery” had taken root in certain branches of woodworking production, owing to the lag in scientific research, I shall cite several examples from recent times. The fact is that some large carriage and furniture factories in the states of New York and Illinois (Chicago) were, almost until the very last moment, firmly convinced that white oak, growing in the south (in the states of Louisiana and the northern part of Texas), possesses greater flexibility and strength than that growing in the north (New York, Illinois, Wisconsin, etc.). And this despite the fact that numerous tests by technological institutes in New York (Columbia University), Boston, Minneapolis, and St. Louis demonstrated the complete identity of the properties of this wood as a construction material. Several years ago chestnut oak was considered unsuitable for ties because of its poor resistance; therefore the logs of this tree, after the removal of the bark from which tannin is produced, were thrown into the forest and rotted there as worthless material. However, testing of this oak for strength by the laboratory of the Forest Department yielded quite satisfactory results, and now the railroads in the state of Alabama, where this oak grows in abundance, use ties almost exclusively of chestnut oak.

These examples sufficiently show how costly the path of guesswork was for American industry where scientific technical investigations were needed; therefore we shall confine ourselves to them. One cannot, however, reproach the Americans for technical conserva—

*) In 1876, at the exhibition in Philadelphia, the organization of workshops and laboratories at the Imperial Moscow Technical School was considered exemplary, and the Boston Technological Institute then borrowed from the Technical School the organization of practical exercises in workshops.

...tism. Since the benefit of a scientific treatment of the matter is obvious to them, they do not spare funds for the perfect organization of the scientific development of one question or another having practical significance.

The extraordinary development of the woodworking industry in the United States over the last 20 years, where the technique of woodworking production now stands immeasurably higher than the European one, prompted the Americans to organize a special laboratory for testing wood as a structural material. This laboratory, established by the Forest Department in Washington, is under the direction of specialist engineers, and the data on wood testing, published annually in the bulletins of the Forest Department, are accepted in the American technical world as a standard, i.e., as the only reliable ones. In addition to this laboratory, mechanical laboratories of the Boston Technological Institute, Washington University in St. Louis, and the University in San Francisco also work for the Forest Department. The general plan of testing was developed by the Washington laboratory.

In the article offered here I wish to acquaint the Russian engineer with the organization of work in this laboratory, where I had occasion to work in the autumn of 1903 and the summer of 1904 on the testing of American longleaf yellow pine (Pinus palustris).

In view of the fact that, of the four laboratories, the laboratory in St. Louis is of the greatest interest as a model testing station, I shall below present its plan and the characteristics of certain testing machines, and now give a standard description of the tree, as a plant species, before it enters the laboratory as the material to be tested.

Questionnaire form for a specimen of wood proposed for testing.

A. General information.

1) Name of the species: (Latin name of the species, scientific term) Pinus Palustris, (market name) yellow pine, (local name) heart of Georgia (state).

2) Place where the tree grew: State of Alabama, Escambia County.

1. Longitude: 86°12′, Latitude: 31°15′. Average elevation above sea level: 100–150 feet. Height of growth of this specimen—125 feet.

4) General character of the locality: mountainous, hilly, level or mixed, etc.

5) Climatic character of the locality: a) temperate, warm, cold, or hot climate; b) mean annual temperature (16°C); c) amount of atmospheric precipitation (50–80 inches of rain).

B. Characteristic of the Soil.

1) Geological formation: deposits.

2) Mineralogical characteristic of the soil: clay, heavy clay, sand, etc.

3) External appearance: bare, covered with grass, dry leaves, etc.

4) Thickness of the chernozem layer, if there is one.

5) Structure, density, and degree of homogeneity of the soil.

6) Moisture: damp, swampy, dry, etc. (Percentage moisture content, if possible).

7) Depth of the subsoil and its nature.

C. Characteristic of the forest from which the specimen under test was taken.

1) Virgin forest or secondary growth?

2) Mixed or homogeneous (Number of species)?

3) If a mixed forest, then what is the predominant species.

4) Continuous stand or clearings? Natural clearings or the result of fires?

5) Average distance between trees (density of the forest).

6) Average age, height, and diameter of the timber species.

Let us consider what has caused such an abundance of questions in groups A, B, and C. First of all, we note that the laboratory, in satisfying the demands of industrial technology, works chiefly on market species. But, striving at the same time to give scientific generalizations to its experiments, it must above all present a complete characterization of the species being tested in connection with the climatic and soil conditions that undoubtedly influence the properties of one and the same species. Therefore, the laboratory does not take wood samples for testing from timber yards, since in that case the laboratory knows only the place of logging and the market name of the wood—data whose accuracy cannot be guaranteed. It is therefore necessary to determine the time of felling (which is almost impossible) and the species, because quite often even experienced lumbermen confuse

different species of trees of one and the same family, differing little in external appearance, but having different properties as building material. This will become clear if one points to the circumstance that in the United States there are counted 14 building species of oak (more than 30 species in all), 13 species of pine, 6 species of ash, and as many of maple, birch, etc.

Having set itself the goal of investigating a definite species of tree, the laboratory first obtains directly from the logging sites several specimens of this species, filling out on the spot the questions of groups A, B, and C; and only then does it proceed to the experiments.

Before turning to the description of my experiments, I consider it necessary to indicate what requirements laboratory work on testing wood must satisfy. The general problem is as follows: it is required to establish the strength of wood, as a building material, in connection with its anatomical structure and chemical composition. In other words, we must answer the following questions:

1) What are the distinguishing features (anatomical or chemical peculiarities) of one or another species of tree, and by what circumstances (growth, climate, etc.) are these peculiarities caused?

2) How do age, rapidity of growth (of one and the same species, depending on climate and soil conditions), and time of felling affect the strength of wood?

3) What is the influence of the anatomical structure of wood on its strength?

4) The influence of specific gravity on strength.

5) The influence of climatic and soil conditions on the strength of wood.

6) The influence of tapping (extraction of resin and turpentine from certain coniferous species) on the strength of wood.

The answers to the questions posed fully determine the properties of wood as a building material.

The task is not as simple as it appears at first glance. For the resolution of these questions the Washington testing laboratory under the Forest Department is divided into four divisions, performing strictly defined functions. 1) The division that collects material for testing (expeditions to logging sites); 2) the division of microscopic (in general physical) and chemical investigations of wood; 3) the division of experiments (the laboratory proper); and, finally, 4) the division of scientific generalizations, which includes the investigators of all the first three divisions.

The functions of the first division consist in selecting on the logging site five quite sound specimens of the proposed

to be tested and to answer the questions of points A, B, and C (see p. 3). Four of the felled trees must be average representatives of the species (diameters of average market size), and the fifth—the best, most fully developed tree. They are usually cut at a height of \(1 \frac{1}{2}\)—\(2 \frac{1}{2}\)—3 arshins, and from the lower end a disk 6 inches high is cut off, which serves for determining the age of the tree. The height of the tree is determined from the base to the top. Then the tree is sawn into logs of market size, cutting from the end of each log disks for determining the age of the various parts of the trunk, and also for determining specific gravity. The sawn disks are weighed at once, wrapped in oiled paper, and immediately sent by mail to the laboratory. On the logs the part of the trunk facing south is marked and, if possible, they are immediately transported to the testing station. In the same manner specimens of the species under test are obtained from other logging sites in various states, so that a sufficiently large number of logs is collected, sufficient for general conclusions.

Not always, however, can the Washington laboratory determine by its own efforts the anatomical structure and chemical composition of one species or another; in such a case the tree is sent to university laboratories for botanical investigation *) and to the chemical laboratory of the Department of Agriculture (in Washington) for chemical investigations.

I personally set myself the task of investigating the influence of specific gravity on the strength of yellow pine. From the further exposition it will be seen that I resolved the problem posed incompletely, having investigated—

a) the influence on strength of the distance from the base of the tree to the part of it being tested.

b) The influence of moisture on the strength of the wood, since moisture in turn affects specific gravity.

c) The influence of the distance of the tested block from the axis of the trunk to the periphery, because, as will be indicated below, specific gravity changes depending on the stated distance.

In what follows I shall set forth how I resolved the problem posed, although not exhausted by points a), b), and c), yet sufficiently embracing the fundamental connection between strength and specific gravity.

*) In America these works are conducted by Professor Filibert Roth—University of Michigan, in the city of Ann Arbor.

Upon receiving the split disks in the laboratory, the first step is to determine the specific gravity of the wood. How cautious the laboratory is in this regard may be judged from the percentage of rejected specimens received. Thus, we had to reject from 10–20% of the shipments solely because in some places the oiled paper had been torn and it was possible to suppose that the wood had lost moisture during transit.

To determine the specific gravity of the wood, the procedure is as follows: first the disk is split into northern and southern parts. Then, from these halves, wedge-shaped or prismatic pieces are split out, measuring 3–5 centimeters along the radius (for hardwoods up to 8 cm is allowed). The accompanying drawing (Fig. 1) explains what has been said.

Splitting with a saw is avoided, because the loss of moisture from the wood (during the experiment it proceeds more vigorously from torn cells adjoining the plane of the cut, whereas in splitting the cells are not torn but merely separated) would be greater. Moreover, when the piece is immersed in water to determine its volume, the cut surface absorbs water into the incised cells, the air-conducting vessels (tracheids), and the empty resin ducts (in coniferous species) arranged along the trunk, thereby reducing the true volume of the wood.

Fig. 1.

The split-out piece is weighed on scales giving an accuracy of up to 0.001 gram for pieces not exceeding 100 grams. When the weight of the freshly cut wood has been determined (a disk delivered with such precautions to the laboratory may be considered part of freshly cut wood), the volume is determined. For this purpose the apparatus shown in Fig. 2 is used.

The metal vessel A (usually iron) has plate-like lugs E and F fastened to it, projecting to the same height \(a\) above its edges.

Distilled or ordinary water is poured into A, the temperature and density of which (the latter for non-distilled water) are determined during the experiment. Into this vessel is inserted a hollow cup B with crossbar CD; the crossbar serves on one side as a handle, and on the other—its chief purpose—it rests

on the lugs E and F, which makes it possible always to bring it up to the level XY, lying 4–6 mm below the level M. The ring-shaped tap is made in order to give the air a free outlet and thus to avoid splashing of water when lowering cup B. The apparatus, shown in this form, is ready for tests.

Determination of the volume is carried out as follows: having removed B and holding it over A, so that the drops fall back into A, one places the basin G, holding not more than 1500 cubic centimeters of water, and carefully lowers into A the piece of wood whose volume is to be measured. Then the wood is pressed down with cup B, resting on E and F, and one waits until the last drop falls from the tap into G. The difference between the weights of the empty basin and of the basin with water gives the weight of the displaced water, from which, with corrections for temperature, the volume is determined. Basin G is of stamped iron and usually does not exceed 150 grams, so that the maximum weight of water and vessel does not exceed 1650 grams, which makes it possible to work with fairly sensitive balances.

Fig. 2.

Against the described method of measuring volume there is usually raised the objection that wood absorbs water, and therefore we cannot determine its volume more or less accurately by immersion; for this reason even the geometrical method of determining volume is preferred. Without mentioning that preparing pieces of wood convenient for geometrical calculations is far from easy when there is a large number of experiments, I shall cite figures characterizing the immersion method and the geometrical one.

From the disk a parallelepiped was sawed out and carefully worked, first on a good planing machine and then by hand, with such

by calculation, so that its transverse dimensions would not exceed 45 mm. Upon micrometric measurement the following was found:

Dimensions were to be: $a_1=a_2=a_3=a_4=a_5=45$ mm	Obtained by measurement	Obtained by measurement	Obtained by measurement
	End I, mm	Middle $a_5$, mm	End II, mm
Width $a_1$ and $a_3$ (top of the bar)	45.28	45.20	45.04
Width $a_1$ and $a_3$ (bottom of the bar)	44.80	44.76	44.64
Height $a_2$ and $a_4$ (right side)	45.38	45.30	45.24
Height $a_2$ and $a_4$ (left side)	45.24	45.38	45.08
Mean	$a_1=a_3=44.95$ mm	$a_2=a_4=45.27$ mm

The smallest error obtained was 0.04 millimeter and the largest 0.38 mm. With an average length of 153.80 mm, a width of 44.95 and a height of 45.27 mm, the volume came out to be 312.965 cubic centimeters. The volume obtained by immersion was 309.700 cubic centimeters, which gave a difference of

Fig. 3.

3.265 cubic centimeters. But if we were to regard as ideally correct the dimensions of the width and height, assumed to be 45 mm, then the volume would be 311.44 cubic centimeters ($45 \times 45 \times 153.8$), i.e., the difference would be 1.74 cubic centimeters, or 0.55%. The question arises whether the volume difference of 3.265 cubic centimeters should be regarded as the result of the absorption of water by the wood through capillary vessels, or, conversely, as a correction for the inaccuracy of the geometrical method of measurement. To resolve this question, before immersing the bar, measurements were made at various points along its length and quite different dimensions were obtained, which in calculation gave a volume differing from the preceding one by 1.1–0.29 cubic centimeters in the positive or negative direction. The difference between the volumes $15.380 \times 4.495 \times 4.525$ and $15.380 \times 4.50 \times 4.50$ gives us an error of 1.52 cubic centimeters for the geometrical method, as compared with the ideally correct dimensions. Let us now compare the error that occurred from the absorption of water by the capillary vessels of the wood.

After one minute of immersion the wood being tested was taken out and, after allowing the last drops of water to fall from it, the sides were lightly wiped, without touching the end surfaces; then it was weighed.

The gain proved to be 3.37 grams. But absorption in the radial and tangential directions almost did not occur over so short an interval of time; therefore, assuming that the water covers, by adhesion, the planed parts of the wood with a layer 0.01 millimeter thick, we find that on this surface \((4 \times 15.38 \times 4.5 = 276.84\ \text{sq. cm})\) there is \(276.84 \times 0.01 = 2.768\ \text{cu. cm}\), or grams, of water. Thus, through the surfaces of saw cuts I and II, \(3.370 - 2.768 = 0.608\) gram, or cubic centimeter, of water is absorbed. Consequently, the loss of volume when determining it by immersion will be 0.608 cubic centimeter, i.e., the error will be almost twice as small as with the most careful measurement of the dimensions and geometrical calculation.

Not to mention the fact that, with a large number of tests, the preparation of such blocks would take a great deal of time and would not give sufficiently accurate results, we would not obtain the correct specific weight of the various parts of the tree trunk also because the sawn block is heterogeneous in density; whereas, by cutting out small pieces in the radial and tangential directions, we have the possibility of obtaining a more or less homogeneous density.

Having determined the specific weight of freshly felled wood, we determined in the same way the specific weight of wood dried to a water content of 15–18%, and it should be noted that the laboratory chiefly practices artificial drying.

To determine the water content in wood, a small oven with double walls is used, the spaces between which are heated by gas firing. To avoid heat losses, the outer walls of the oven are covered with asbestos sheathing, which makes it possible to keep the temperature in the oven with fluctuations of up to \(1^\circ\)–\(1 \tfrac{1}{2}^\circ\) C.

Wood specimens are prepared in two ways: 1) from freshly felled wood, shavings are bored out from the periphery to a depth of \(1/3\) of the diameter, the boreholes being made one from another at a distance of 20 feet; and 2) from various parts of the disk, thin sticks (no thicker than matches) are split off. After preliminary weighing, the specimens thus obtained are placed in the oven, the temperature of which is gradually raised from 30 to \(105^\circ\) Celsius. After twenty-four hours of drying all the water evaporates, and the difference in weights determines the water content in the wood. In determining the percentage water content, Bauschinger allows¹ a temperature of \(105^\circ\) C., and he observed an increase

the absolute weight of the wood (in resinous coniferous species) from a certain minimum by 0.2% and took the smallest weight. The phenomenon of an increase in weight, for which he gives no explanation, probably occurs from the oxidation of resins². In order to establish this phenomenon and the proposed explanation of it, it is necessary to dry the wood in some neutral gas (nitrogen or carbon dioxide). Part of this question was investigated by me in the laboratory of Prof. N. M. Kizhner, namely the question of the increase in weight when pine is dried at a temperature of 100° and then 105° C.

For the experiments I took pine sawdust from a tree felled a year earlier (\(P_1\) and \(P_2\)), sawdust from a log five years old (\(P_3\)), and pine resin (\(P_4\)) that had exuded from the bark and had already partly oxidized at room-air temperature.

The results of weighings, with an accuracy to 0.1 milligram, were as follows:—

	Initial weight	Minimum at 100°C, interval between weighings 3 h	Minimum at 100°C, interval between weighings 3 h	At 105°C. Intermediate time 18 h	At 105°C. Intermediate time 18 h
\(P_1\)	3.6388 grm.	3.0355 grm.	3.0355 grm.	2.9955 grm.	2.9963 grm.
\(P_2\)	3.7068 ”	3.3420 ”	3.3420 ”	3.3363 ”	3.3363 ”
\(P_3\)	4.3838 ”	3.8170 ”	3.8170 ”	3.8108 ”	3.8103 ”
\(P_4\)	1.2275 ”	1.2114 ”	1.2100 ”	1.1953 ”	1.1920 ”

In the interval between the 1st and 2nd weighings (according to the table) there was one further weighing. Between the 1st and 2nd weighings (according to the table) 4–5 hours elapsed; between the 3rd and 4th, 2–3 hours. Thus, only for \(P_1\) was an increase in weight obtained at a temperature of 105°C, and that only by 0.02%; in the other cases there was a loss of weight. But what is important is precisely the fall in weight from its known stationary state, when, on the basis of two adjacent weighings (the 2nd and 3rd columns), we have the right to conclude that the sawdust no longer contains water. Obviously, the loss of weight is due to the evaporation of resinous substances, of which \(P_4\) convinces us. At the end of the experiment the sawdust was completely without a resinous odor. These experiments provide full grounds for recommending that the drying of wood, for determining the percentage of water, be carried out at a temperature not above 100°C.

The results of determinations of the water content and specific gravity from 45 specimens of yellow pine give the following table:—

	Specific gravity	% water content	At 15–18% water
Lower part of the trunk	0.60–1.19	48–52%	0.45–1.04
Middle . . . . .	0.71–0.98	48–52%	0.56–0.86
Upper . . . . .	0.68–1.09	48–52%	0.48–0.91

Having determined the specific gravity and the moisture of the wood, we proceeded to experiments on shrinkage. If \(V\) is the volume of freshly felled wood, \(v\) the volume of dry wood, then the shrinkage is expressed by:

\[ \frac{V-v}{V}. \]

Strictly speaking, these tests are not important for wood as a building material, since, in studying the strength of wood, we must consider it as a permanent material, and it is such at a moisture content of 12–18 percent. But for America these investigations have market significance. The matter is that among Americans there exist a considerable number (up to 43) of empirical determinations of the volume of wood by its greatest diameter. They determine the volume starting from the conventional unit—board measure \((1'' \times 1' \times 1')\).

But sawn market timber, containing up to 30% moisture after 4–6 months from the time of sawing, gives a smaller volume. Therefore sawmills require a certain percentage deduction from logging camps for shrinkage. This circumstance caused endless disputes between the owners of logging camps and sawmills, which, properly speaking, prompted the Washington laboratory to carry out tests on shrinkage, in order to work out a more or less definite coefficient.

The experiments were continued with the aid of an instrument used for determining specific gravity, with the moisture of the tested disks being reduced to 15%–30%.

Summarizing the results of the investigation of wood for moisture, specific gravity, and shrinkage, I shall present this in a diagram, where it is easiest to show it as a function of the age of the tree (the number of annual layers) at various points along the height and with the orientation of the trunk to the south or north. I shall give here tree No. 1 in the order of testing, and shall limit myself to only one specimen, since the law of dependence of moisture, specific gravity, and shrinkage on age and on the orientation of the trunk to the south or north is sufficiently determined on a single specimen.

Pinus palustris No. 1.

Age of the tree—182 years.

Height to the first branch—70 feet.

Total height—113 feet.

Diameter at the base—19 inches.

In the appended diagram the designations are as follows:

A—the specific gravity of the wood at 12–15% water (artificially dried).

B—the percentage of water in freshly felled wood.

C—shrinkage, in percent, relative to the volume of freshly felled wood.

N—the number of annual layers (age).

I, II, III, IV, etc.—disks 4″–6″ in height, from which these data were obtained.

The right half of the trunk faces north, and the left half faces south.

Diagram No. 1.
Pinus palustris No. 1.

Fig. 4

Examining the figures of this diagram, we observe that the specific gravity of the wood decreases as the tested samples approach the top, just as does the percentage content of water. The percentage of shrinkage also decreases as the top is approached. At first glance this phenomenon seems strange. It would seem that the more moist part of the tree should give a greater percentage of shrinkage, whereas here it is just the opposite. The only explanation that can be given for this phenomenon is that the younger cells of the tree (nearer the top) resist the loss of moisture more vigorously than the older ones, not shriveling and preserving their original form. And indeed, the younger parts of the trunk, when dried under the same conditions of temperature and time as the older parts, give a greater percentage of cracks.

Comparing then the corresponding figures for the northern and southern parts of the trunk, we observe that the specific gravity, the water content and

drying of the first is lower than that of the second. In other words, the vitality of the southern part of the trunk proceeds more vigorously than that of the northern. This is also understandable, since the southern part of the trunk is exposed to the action of the sun or,—if the sun does not penetrate into dense stands,—to the action of the warm southern wind and is protected from the cold; whereas the northern part is in worse conditions of existence. This circumstance is important in that it indicates a difference in the strength of the southern and northern parts of the trunk as a function of specific gravity, which is important to bear in mind when organizing tests of wood.

No. 2.

Fig. 5.

From diagram No. 1 we observe that:

a) the specific gravity falls in the direction toward the top with increasing moisture content of the wood.

The latter depends, of course, on the lesser compactness of the cells at the top. But when studying the wood in the small section between disks I and II, where the percentage of water content changes hardly at all, a phenomenon of direct dependence is observed:—

b) the specific gravity decreases with the decrease (during drying of the wood) in the percentage water content, which is clearly seen from the diagram, resulting from 16 determinations at different water contents. In other words, with a decrease in water content from 43% to 8%, the specific gravity decreased from 0.73 to 0.546.

Further, investigations of the percentage water content of the wood and of the specific gravity (the objects of the experiments were the same disks I, II, etc.), in the direction from the center to the periphery, showed that both these quantities increase (diagram No. 3).

Diagram No. 3 gives:

c) the curve AB—the change in specific gravity, and CD—the percentage water content as a function of the distance from the axis of the trunk.

Thus, these preliminary investigations, summarized in diagrams Nos. 1, 2, and 3, fully determine the character of our investigations. It is clear that the strength of the wood will vary both from the base to the top, and from the axis to the periphery.

iphery; moreover, parts of the trunk facing north must differ in strength from the southern parts of the corresponding height sections.

Diagram No. 3.

Fig. 6.

Before proceeding to the further description of the tests, I consider it not superfluous to give a characterization of the species being tested as a botanical specimen.

The species “Pinus palustris” occurs almost exclusively in the south of the United States (Florida, Georgia, Louisiana, and along the Mississippi River valley), in low-lying places—from 60 to 300–400 feet above sea level. Its favorite soil is alluvial deposits; the soil is predominantly moist. In the general appearance of this pine, hanging tassels of a peculiar moss are noticeable, affecting almost all tree species of the south of the country, not only conifers, including the cypress tree, but also certain species of oak. From the figures appended to the text one may judge both the structure of the wood and its development from the seed. Detailed explanations are given beneath the figures in the appendix.

Preparation for the Experiments

We have already established the variability of the moisture content and specific gravity of the wood both from the base to the top and from the axis to the periphery, as well as in dependence on which way the trunk is turned—toward the north or toward the south. In connection with this, the tests must also be carried out accordingly. For the tests, the corresponding beams are sawn from the trunk; these are dressed on longitudinal planing machines, which work one or another small beam simultaneously from four sides. Of course, it is not possible to obtain the desired size of the cross-sectional area exactly, but the error is, comparatively, insignif—

— up to 1–2 mm in linear dimensions. The bars are sawn out as shown in Figures I to VIII.

Fig. I.  Fig. II.  Fig. III.  Fig. IV.  Fig. V.

Fig. VI.  Fig. VII.  Fig. VIII.

Figures I, II, III, and V show the sawn bars subject to testing, the purpose of which is to establish the relationship between the distance from the base of the tree to the bar and its strength. The same arrangement of cuts is also used for two adjacent sections of the log when determining the relationship between strength and moisture content. Figure IV gives the form of the cut for testing full-sized columns—bars with a cross-sectional area of up to \(12'' \times 14''\) and a length of up to 30 feet. Figures VI and VII show the arrangement of the cuts depending on whether the trunk faces north or south. And, finally, Fig. VIII gives the bars for determining the relationship of strength to their distance from the axis of the trunk. The number of bars in Fig. VIII depends, of course, on the magnitude of the tree diameter, and the drawing shows the minimum number of them, taking into account the bending test, which requires the smallest cross-sectional area of the bar, \(75 \text{ mm} \times 75 \text{ mm}\).

Bending tests. The usual size of bars tested in bending, adopted by the Washington Laboratory, is \(3\frac{1}{2}—4 \times 3\frac{1}{2}—4\) or \(4 \times 8\) inches (English) in cross-sectional area and from 60 to 120 inches in length, using for this purpose a Riehlé universal machine rated at 50,000

Seed Germination

e—the first period of seed germination.
a—the first sprout in early spring with 8 leaves.
b—age 2–3 weeks.
c—at the end of the first or the beginning of the second season.
d—age 3–4 years.

Spring branch of Pinus palustris.

a—end of a branch with a spring shoot.
f—tuft of (three) needles—normal appearance of the leaves.
h and g—needle sheath enlarged 3 and 9 times.
e—section of a needle socket (linear enlargement 30 times).
d—section of a young needle (enlarged 30 times).
c—section of a mature needle (enlarged 45 times).
b—longitudinal section of a needle (enlarged 45 times).

a—branch with male flowers.
b—branch with female flowers and a young cone. Flowering time: in the twenties of March.
c and d—separated flower catkin.
e, f, and g—stamens of the flower.
h—female flower.
k, l, and m—scales at the base of the female flower.
n and p—scales with ovules.

a—mature open cone.
b and c, d and e—seeds with wings; from the first two the wings have been detached.
f and h—separate cone scales; front and rear views.

a—appearance of the bark of yellow pine.
b—radial section; c—tangential section, made somewhat at an angle to the axis of the tree.
d—end section.

Specimens of the tested bars.

A and B—axial compression; C—shearing; D—tension.
The lower number denotes the number of pounds of ultimate load.

kg., which will be discussed below. For beams of size \(8'' \times 14\)—\(16'' \times 16\)—18 they use the Riehle machine for 100,000 kg. The test is carried out to the point of fracture, and for each beam a diagram is constructed. To avoid crushing at the points of support and pressure (in the middle of the beam), pads are provided, as shown in the drawing (Fig. 7).

Fig. 7.

In my bending tests I calculated only the elastic and breaking load per 1 sq. centimeter and the modulus of elasticity; but the Washington laboratory proposes more data for characterizing the strength of the material. These data are as follows:—

1) First of all, the load per 1 sq. centimeter within the elastic limit:—

\[ K_b=\frac{Pl}{4W}, \quad \text{or} \quad \frac{3}{2}\frac{Pl}{bh^2}, \]

(\(W\)—the section modulus for a rectangular beam).

2) The load per 1 sq. centimeter at the breaking force (modulus of rupture):—

\[ K'_b=\frac{P'l}{4W}=\frac{3}{2}\frac{P'l}{bh^2} \]

Here \(P'\) denotes the force at the moment of rupture. Of course, one cannot assume the same law of variation of the load beyond the elastic limit as we assume within its limits; nevertheless this

an arbitrary and conventional formula for \(K'\), characterizes, to a certain degree, the resistance of the material at the moment of rupture.

3) The greatest shearing stresses:—

\[ K_s = \frac{3}{2}\frac{P}{bh} \]

4) Modulus of elasticity:—

\[ E = \frac{P}{f}\frac{l^3}{4bh^3}, \]

where \(f\) is the deflection.

5) Work of the load up to the elastic limit (elastic work—elastic resilience \(R\)):—

\[ \text{Elastic work} — R=\frac{P \cdot f}{2}. \]

6) And, finally, the work of the load up to the elastic limit per 1 cubic inch (or centimeter) \(r\):—

\[ \text{Elastic work per 1 sq. cm.}=\frac{Pf}{2w}, \]

where \(w\) is the volume of the tested bar.

Sometimes the total work of the load up to the moment of rupture is calculated (both total and per unit of volume), for which, consequently, it is necessary to calculate the area \(ABCD\) of the diagram (diagram No. 4) and divide by the volume of the tested bar.

Test in tension. In testing in tension it is recommended to take the cross-section not round, as is usually customary among us, but rectangular, in view of the fact that in lathe-working the bar there occurs tearing up and cutting of the outer fibers, and this reduces the strength of the specimen tested. Although a round cross-section is also used. For testing, the same Riehle or Olsen machine is employed. The fastening and the form of the bar are shown in Fig. 2.

Compression along the fibers. The dimensions of the bar (usually of square cross-section) are \(3—4 \times 3—4\) inches in area and \(7—12\) inches in length (Fig. 4). But in the laboratory of Washington University in St. Louis, compression tests are made on timbers of colossal dimensions, as, for example, with an area of \(12 \times 14\) inches and a length up to 36 feet, by a special machine developing up to 500,000 kilograms.

Compression perpendicular to the direction of the fibers. The area of compression of the tested bar is taken up to 16 square inches. For these

Fig. 2.

A — B

Section along AB
I

Section along AB
II

Tension.

Fig. 3.

a
b
c
d

Shearing.

Fig. 4.

Axial compression.

Fig. 5.

40 × 60

$f_1$
$f_2$

Compression ⟂ to the direction of the fibers.

machines of Riehle or Olsen serve for the tests. The placement of the timber is shown in Fig. 5.

Shearing. For shearing tests, special timbers up to 10 inches long are prepared, with rectangular holes bored (by a mortising machine) at both ends at a distance of 1–1½ inches from the ends, and with one hole in the middle. The shearing area is 1–2 square inches, and the direction of shearing is taken, for one and the same timber, as radial and tangential.

From the appended drawings the described method of testing is sufficiently clear (Fig. 3), but it cannot at all be recommended as rational. The point is that the small sleeve with screw a is installed in order to keep the shearing ends of the timber from being pushed apart by the fork at the moment of shearing, since bending of these ends is assumed. But, first, there is no reason to suppose bending here; and, second, the clamping of the ends produces compressive forces, which increases the resistance to shearing.

Unfortunately, for lack of time, I was unable to carry out torsion tests. It must be said, however, that the laboratory, too, almost does not perform these tests, in view of the rare cases in which timber is used in structures where it is subjected to torsion.

II.

Study of the strength of wood as a function of the distance of the log from the base.

The tested specimen of yellow pine (yellow pine, long-leaf pine)—Pinus palustris—was obtained from the state of Alabama, Escambia County.

A) Longitude 86°,25′; latitude 31°,10′. The average elevation of the locality above sea level is from 75 to 100 feet. The locality is hilly; mean annual temperature 14° C.; annual atmospheric precipitation—100–120 cm.

B) Geological structure of the locality—alluvial deposits; soil—clayey sand, well moistened; soil color—ash-gray.

C) Stand moderately dense; forest virgin, continuous yellow pine; average height of the merchantable forest—90 feet, diameter—18½ inches.

Tested specimen. Tree No. 1. Diameter at breast height—19 inches. Total height 113 feet. Age according to annual rings at a height of 18 inches from the root (stump height)—182 years. Distance from the nearest trees 20–50 feet.

The tree was sawn into seven logs, with disks cut out for determining the specific gravity and moisture content. Table No. 1 below characterizes the distance of the log from the root, its length, and its mean diameter.

Nos. of logs and disks	Distance from plane of stump	Length of log	Mean diameter
I	0′8″	12′4″	19″
II	13′8″	5′4″	15,25″
III	19′8	12′4″	14⅜″
IV	32′8″	14′4″	13¾″
V	47′8″	9′4″	13¼″
VI	57′8	9′4″	12¼″
VII	67′8	9′4″	11½″

Table No. 1.

The tests were carried out, as may be seen from the summary table, at a reduced percentage content of water, except for beam No. 1. The drying of the wood was artificial, and the water content was reduced to the limits of the usual moisture of wood intended for work.

Before speaking of the general conclusions of the test, let us consider separately the character of the dependence of the load and the deflection for bending, i.e., the bending diagrams, and also the diagrams of compression along and across the fibers. The diagrams appended below are typical; therefore I shall limit myself to only a few specimens for each category of tests.

It should first be noted that the Washington laboratory builds diagrams for bending and compression of both kinds invariably for each beam, since it is considered important to determine the work of bending or compression per unit volume (per cubic inch), which serves as an additional datum for determining the strength of one or another species of wood.

For bending I give here three diagrams:—1) for a full beam (Fig. IV); 2) for a beam of the type of Figs. II and III, No. 1, and 3) for a beam of the type of Fig. VII, Nos. 1, 5, 9, or 12.

Beam of the 1st type (No. 62). Tested on a Riehle machine at 100,000 kg. The speed of the machine during the test was 3.8 mm per minute.

Dimensions of the beam	b = 250 mm.
Dimensions of the beam	h = 360 mm.
Dimensions of the beam	l = 4720 mm.
Moisture content	37.3%
Weight of the whole beam	238 kilograms.
Specific gravity	0.64
Specific gravity at 15% moisture	0.47

The results of the test, as may be seen from the diagram constructed from the records, are as follows:

Limit of elastic load—9,000 kilograms.

Breaking load—16,800 k.

Deflection within the limits of elasticity—21 mm.

Beginning of test	Load, kg	Deflection arrow, in.
10 h. 8 m.	0	0
10 h. 8 m.	500	0.04
End of test	1000	0.085
End of test	1500	0.13
10 h. 47 m.	2000	0.18
Duration, 35 m.	2500	0.23
Duration, 35 m.	3000	0.28
Duration, 35 m.	3500	0.33
Duration, 35 m.	4000	0.37
Duration, 35 m.	4500	0.42
Duration, 35 m.	5000	0.46
Duration, 35 m.	5500	0.51
Duration, 35 m.	6000	0.555
Duration, 35 m.	6500	0.605
Duration, 35 m.	7000	0.65
Duration, 35 m.	7500	0.71
Duration, 35 m.	8000	0.755
Duration, 35 m.	8500	0.80
Duration, 35 m.	9000	0.85
Duration, 35 m.	9500	0.90
Duration, 35 m.	10000	0.95
Duration, 35 m.	10500	1.005
Duration, 35 m.	11000	1.065
Duration, 35 m.	11500	1.12
Duration, 35 m.	12000	1.18
Duration, 35 m.	12500	1.25
Duration, 35 m.	13000	1.32
Duration, 35 m.	13500	1.39
Duration, 35 m.	14000	1.47
Duration, 35 m.	14500	1.56
Duration, 35 m.	15000	1.67
Duration, 35 m.	15500	1.815
Duration, 35 m.	16000	1.985
Duration, 35 m.	16500	2.22
Duration, 35 m.	18000	2.50

Table No. 2.

The deflection arrow for the ultimate load was 63.5 mm. Below I give one of the tables of notes made during the test.

As is seen from the appended table (Table No. 2), the diagram was plotted by loads of 1000 kg, i.e., through one point, because of the small scale of the diagram itself, where 0.2 inch of deflection corresponds to \(1/2\) inch of scale, and a load of 1000 kg corresponds to \(1/4\) inch.

The corresponding results of the calculation are as follows:

\[ K_b = 216 \text{ kg per sq. cm.} \]

\[ K'_b = 401 \quad ” \quad ” \quad ” \quad ” \]

\[ K_s = 31.0 \quad ” \quad ” \quad ” \quad ” \]

\[ E = 101000 \quad ” \quad ” \quad ” \quad ” \]

\[ \text{Elastic work} = \frac{AB' \times BB'}{2} = 9000 \text{ kg-cm.} \]

\[ ” \quad ” \text{ per } 1 \text{ cubic cm.} = 0.026 \text{ kg-cm.} \]

Beam of the 2nd type (No. 52-B-1).

Dimensions of the beam	\(b = 99\) mm.
Dimensions of the beam	\(h = 200\) mm.
Dimensions of the beam	\(l = 2000\) mm.
Moisture content	34.2%.
Weight of beam	24.5 kg.
Specific gravity	0.57.
Specific gravity at 15% water	0.425.

Tested on a Riehlé machine at 100000 kg.

Speed of the machine—3.8 mm per minute.

Limit of elastic load—4250 kg.

Breaking load—7000 kg.

Deflection arrow within the limits of elasticity—13 mm.

Deflection arrow for the ultimate load—28.5 mm.

Finally, a beam of the 3rd type (No. 62-A 2).

Dimensions of the beam:

\[ \begin{aligned} b &= 50\ \text{mm.}\\ h &= 50\ \text{mm.}\\ l &= 762\ \text{mm.} \end{aligned} \]

Moisture content—35.2%.

Weight of the beam—1.51 kilograms.

Specific gravity—0.601.

“ at 15% water—0.39.

Tested on an Olsen machine of 15,000 kilograms.

Speed of the machine—2.9 mm per minute.

Limit of elastic load—400 kilograms.

Breaking load—730 kilograms.

Deflection arrow within the limits of elasticity—6.35 mm.

Deflection arrow for the limiting load—34.2 mm.

Since the diagrams for beams of the 2nd and 3rd type are completely constructed, I shall not give the tables of load readings and deflection arrows.

The results of the tests of these three beams are summarized in Table No. 3:

	Beam No. 62.	Beam No. 62. B—1.	Beam No. 62. A—2.
\(K_b=\)	228 kilograms	272 kilograms	276 kilograms
\(K'_b=\)	458 ”	472 ”	338 ”
\(K_s=\)	31.0 ”	48.2 ”	35.8 ”
\(E=\)	101000	88100	90000
Elastic work \(=\)	9000 kilogram-centimeters	2762 kilogram-centimeters	127 kilogram-centimeters
” per 1 cubic inch \(=\)	0.026 ” ”	0.065 ” ”	0.067 ” ”

Table No. 3.

Comparing the strength data of these three beams, we note that the strongest one (elastic load per square inch and elastic work) is the beam of the third type. Of course, to draw a general conclusion in this sense on the basis of a single experiment is too risky; but a whole series of experiments of the Washington laboratory on other species of wood, as well as my tests on Pinus palustris, have shown that this is indeed so. This conclusion appears unexpected and, at first glance, paradoxical, since it would seem that the beam of the 3rd type ought to give less resistance per unit area, as less

No. 4.

Beam No. 62.

BENDING DIAGRAM

Vertical axis: Load, in kilograms

Scale markings: 1000; 2000; 5000; 10000; 15000; 20000.

Horizontal axis: Deflection, in inches

Scale markings: 0; 0.1; 0.2; 0.5; 1.0; 1.5; 2.0; 2.5.

Curve label: Theoretical parabola.

No. 5

STRESS DIAGRAM.

Beam No. 62-B-1.

No. 6

STRESS DIAGRAM.

Beam No. 62-A-2.

dense (lying closer to the periphery of the tree). Nevertheless, here the question comes down not so much to the structure of the scantling as to its size. It is clear that a scantling of the 1st and 2nd type, representing cuts from a whole log or from half a log, may have hidden defects: shakes (cracks), rot damage, overgrown knots or cavities, and so forth—which cannot be discovered in a full-sized scantling by external inspection. This, of course, lowers the strength of the wood. When, however, we are dealing with a scantling of small cross-section, such as the 3rd type, we can almost always select a completely sound and undamaged specimen. It goes without saying that small-sized scantlings located nearer to the axis of the trunk are stronger, as will be shown below by a series of tests. For the present I shall only note that full-sized scantlings are 10–20% weaker than small ones.

Axial Compression.

Let us now consider axial compression (along the fibers).

We have two scantlings: No. 1 and No. 2.

	No. 1	No. 2
Dimensions of the scantling.	a = 100 mm. b = 200 mm. h = 457 mm.	a = 100 mm. b = 100 mm. h = 457 mm.
Moisture.	31.9%	34.2%
Weight.	6.2 kg	6.0 kg
Specific weight at 15%.	0.42	0.45
Elastic limit of load.	42500 kg	38500 kg
Breaking load.	53500 kg	53000 kg
\(\lambda = A_1B_1\)	0.72 mm	0.89 mm
\(\lambda_1 = A_1C_1\)	1.47 mm	2.36 mm

Both scantlings were tested on a Riehle 100,000-kilogram machine. The speed of the machine was 3.8 mm per minute. \(\lambda\) is the magnitude of the proportional compression; \(\lambda_1\) is the magnitude of the compression (total) up to the moment of failure.

As is seen from the designations \(\lambda\) and \(\lambda_1\), equal to \(A_1B_1\) and \(A_1C_1\) (see the diagrams), I count the portion of proportional compression only from \(AB\), which requires explanation.

First of all, I note that the diagrams of axial compression came out entirely normal, although within the load range up to 18,000 kilograms no compression proportional to the loads is observed. This phenomenon is caused, undoubtedly, by the circumstance that the first moments ...

COMPRESSION.

DIAGRAMS OF AXIAL COMPRESSION.

Specimen No. 1.

Diagram text: Specimen No. 1. Axial compression (longitudinal).

Vertical axis: Load in kilograms.
Horizontal axis: Amount of compression in inches.

Visible scale markings:

• Load: 5,000; 10,000; 25,000; 50,000 kg
• Compression: 0; 0.01″; 0.02″; 0.05″; 0.1″

Specimen No. 2.

Diagram text: Specimen No. 2. Axial compression.

Vertical axis: Load in kilograms.
Horizontal axis: Amount of compression in inches.

Visible scale markings:

• Load: 5,000; 10,000; 25,000; 50,000 kg
• Compression: 0; 0.01; 0.02; 0.05; 0.1″; 0.108555″

under load there occurs a crushing of the fibers at the ends of the scantling, where they have been separated by sawing. This phenomenon is observed positively in all species of wood tested in axial compression, with the only difference that for the harder species the portion \(OA\) is smaller. Therefore it is precisely the portion \(ABC\) that must be regarded as the normal process of resistance. Of course, we must take the full load at \(C\), but the law of proportionality can be derived only on the portion \(AB\).

Compression perpendicular to the direction of the fibers.

Scantlings Nos. 1 and 2.

	No. 1	No. 2
Dimensions of the scantling.	\(a = 100\) mm. \(b = 200\) mm. \(h = 200\) mm. \(l = 610\) mm.	\(a = 98\) mm. \(b = 196\) mm. \(h = 198\) mm. \(l = 508\) mm.
Moisture.	35.9%	34.2%
Specific gravity.	0.44	0.33
Breaking load.	9500 kil.	11500 kil.

Both scantlings were tested on the Riehle machine of 100,000 kg. The speed of the machine was 3.8 mm per minute.

The phenomenon of non-proportionality of the compressive load at the beginning of compression, observed in the preceding experiments, is also noted here, but to a lesser degree. Undoubtedly, this phenomenon is caused by the very nature of the experiment, where the middle part of the scantling is subjected to compression, which at first causes the phenomenon of crushing over the area of the steel bearing plate. Moreover, it may be said that here a compound resistance is obtained (Fig. 5)—compression and bending. Therefore, after the first moments of the process, when compression of the surface fibers has occurred, the law of proportionality is observed up to point \(B\).

At point \(B\) the critical moment occurs, and up to \(C\) there is no proportionality. Then, from \(C\) to \(D\), proportionality is observed, the amount of compression per unit load being considerably greater than on the portion up to \(B\). Undoubtedly, with such a character of the experiment, on the portion of proportionality up to \(B\) there must be the phenomenon of compression of the part of the scantling under the bearing plate and bending of its ends, since these ends are subjected to the action of forces \((f_1 \ldots f_n)\), reactions tending to raise them, i.e., to bend the scantling. From \(B\) to \(C\) the proportionality is disturbed; evidently, this part of the diagram characterizes the moment of lifting of the ends of the scantling. And, finally, from \(C\) to the moment of failure, compression occurs without the influence of bending.

Compression perpendicular to the direction of the fibers.

COMPRESSION DIAGRAM

Block No. 1.

Vertical axis: Stress per square inch.
Horizontal axis: Amount of compression in inches.

Scale marks on the horizontal axis: 0; 0.1; 0.2; 0.3; 0.5; 0.8.

Handwritten note in the plot: “Plane No. 1 — side surface.”

Curve points marked: A, B, C, D, E, F.

COMPRESSION DIAGRAM

Block No. 2.

Vertical axis: Stress per square inch.
Horizontal axis: Compression in inches.

Scale marks on the horizontal axis: 0; 0.2; 0.5; 0.7.

Handwritten note in the plot: “Plane No. 2 — [[unclear: handwritten surface designation]] surface.”

Curve points marked: A, B, C, D, E.

Such an arrangement of the experiment, complicating the testing and, it must be said, not providing precise data characterizing compression perpendicular to the direction of the fibers, is explained by the laboratory on the grounds that the character of these experiments corresponds most closely to actual cases of resistance to compression of this kind, and that cases of compression of a beam along its entire length are wholly absent in construction practice.

As for the tension diagrams, it was not necessary to construct them, in view of the fact that it was impossible to determine accurately the elongation of the beam, since crushing and slipping of it in the jaws of the grip prevent a correct record of the elongation.

Let us now proceed to the investigation of the results obtained for the strength of the wood as a function of the distance of the timbers sawn successively from its base to the top. On pages

Bending.

Diagram 7.

20 and 21 a description was given of the specimen and of the dimensions of the logs. I shall confine myself to a single table summarizing the test results (two specimens of yellow pine were tested), since the law of variation of strength is sufficiently clear also from this table.

Table 4-a. Characterizing the strength of wood at various heights.

Log no. and form of cutting from the log	Specimen no.	Moisture, %	Apparent weight	Dimensions of specimen, \(l\), mm	Dimensions of specimen, \(h\), mm	Dimensions of specimen, \(b\), mm	Bending: maximum load, \(P_1\), kg	Bending: \(h=b=\sqrt[3]{4W}\), mm	Bending: breaking load, \(P_1\), kg	Bending: \(h=b=\sqrt[3]{4W}\), mm	Compression along fibers: dimensions, length, mm	Compression along fibers: dimensions, width, mm	Compression along fibers: dimensions, height, mm	Compression along fibers: breaking load per 1 sq. cm	Compression across fibers: dimensions, length, mm	Compression across fibers: dimensions, width, mm	Compression across fibers: breaking load per 1 sq. cm	Tension: area of rupture, sq. cm	Tension: breaking load per 1 sq. cm	Shearing: area of rupture, sq. cm	Shearing: breaking load per 1 sq. cm	Cleavage: area of cleavage, sq. cm	Cleavage: maximum force, kg
I	1*)	4.5	0.98	1524	89.0	94.0	475	—	250	—	203.2	81.2	286.0	92.7	86.4	43.4	5.81	422.0	22.8	25.6	—	—	22.8
I	1*)	20.4	0.785	1524	86.4	85.2	908.0	—	702	—	205.7	73.6	584.6	84.8	86.2	71.2	3.68	824.3	23.2	56.4	—	—	23.2
I	2	13.1	0.767	1254	88.5	94.3	930.5	—	712	—	203.2	82.3	476.5	95.0	94.2	59.7	6.38	1009.7	25.8	63.5	—	—	25.8
I	2	11.8	0.730	1254	92.5	93.0	983.4	—	745	—	203.0	68.1	598.2	94.0	92.3	72.1	4.58	1112.3	26.5	56.5	—	—	26.5
I	3	19.2	0.80	2508	200.0	94.6	976.6	—	846	—	206.0	95.0	486.0	100.2	92.6	66.7	6.44	980.8	26.0	46.5	—	—	26.0
II	1	18.7	0.73	1254	92.0	95.0	908.2	—	580	—	203.2	83.6	450.2	87.9	94.2	68.5	5.81	893.3	29.3	41.5	—	—	29.3
II	2	20.7	0.64	1254	92.0	94.2	890.0	—	624	—	203.2	80.6	449.0	94.5	86.5	70.8	4.20	880.5	25.1	51.5	—	—	25.1
II	3**)	18.9	0.70	1254	92.6	94.2	870.5	—	512	—	203.2	79.8	405.0	94.5	82.0	49.5	5.30	799.4	28.4	45.8	—	—	28.4
III	1	17.1	0.72	1254	88.9	95.6	830	—	563	—	205.7	82.4	478.3	94.7	87.5	61.5	5.53	854.5	24.5	54.3	—	—	24.5
III	1	19.2	0.76	1254	93.5	93.5	920	—	455	—	203.2	74.8	566.0	91.4	92.2	67.0	5.93	892.3	24.2	57.7	—	—	24.2
III	2	16.0	0.70	1254	88.2	94.5	850	—	425	—	203.2	82.2	465.0	94.3	84.4	69.1	6.99	751.0	27.7	40.8	—	—	27.7
III	2	14.4	0.762	1254	93.5	93.5	915	—	845	—	200.6	68.9	584.1	94.5	90.5	70.5	4.50	785.2	25.6	50.7	—	—	25.6
III	3	25.2	0.608	2508	200.1	93.3	1000	—	740	—	200.6	81.1	508.0	86.4	90.1	71.2	5.72	658.8	25.8	56.2	—	—	25.8
IV	1	20.4	0.73	1254	84.0	89.5	843	—	586	—	205.7	82.6	404.0	94.2	85.5	58.5	5.31	645.4	26.5	48.1	—	—	26.5
IV	1	16.1	0.75	1254	87.5	88.0	622	—	715	—	203.2	68.8	492.5	94.3	91.0	71.5	4.53	682.3	25.8	50.6	—	—	25.8
IV	2	21.0	0.700	1254	82.1	89.2	775	—	322	—	205.7	82.1	407.5	94.8	83.1	58.2	6.55	620.5	25.2	50.0	—	—	25.2
IV	2	13.9	0.762	1254	84.0	82.3	846	—	625	—	205.7	77.0	408.0	87.2	84.5	71.0	3.75	624.3	25.5	56.3	—	—	25.5
VI	1	18.1	0.64	1254	88.9	94.0	710.0	—	522	—	205.3	82.6	451.0	94.5	88.5	70.0	6.08	634.5	25.4	58.5	—	—	25.4
VI	2	24.2	0.74	1254	89.8	94.3	663.0	—	494	—	205.7	85.1	429.8	89.4	94.6	71.3	7.05	598.6	26.7	50.1	—	—	26.7
VII	1	18.5	0.68	1254	87.9	95.2	620.4	—	465	—	203.0	81.8	394	94.7	87.5	71.2	6.53	590.5	26.0	41.6	—	—	26.0
VII	2	17.7	0.82	1254	89.2	94.5	985.0	—	415	—	205.7	84.5	345.0	94.5	89.4	69.5	6.03	607.8	26.7	49.5	—	—	26.7

*) Logs I, II, III and IV were clamped at a length of 1-to and 2-to.

**) Specimen no. 2-to showed cracking; the log was broken into V-shaped pieces.

In the tested specimen it was necessary to discard log VII, since it proved to be defective.

In preparing the diagrams showing the relation between the distance from the base of the tree and the data characterizing strength, we take only those beams that contain approximately the same percentage of water.

These beams are as follows: from log I, beams 1 and 3; from log II—1, 2, and 3; from log III—1, 1, and 3; from log IV—1, 1, and 2; from log VI—1 and 2; from log VII—1 and 2. In all, 15 beams.

Let us first consider bending. From diagram No. 7, constructed according to the data of the table, one clearly sees the decrease in the strength of the wood as the cut of the beam is removed from the base of the trunk. The elastic load falls from 846 kilograms to 415, i.e., in the present case it decreases by more than half.

Bending.

Modulus of rupture (kilograms per 1 sq. cm).

Distance from the ground (in feet).

Diagram 8.

Compression along the fibers.

Breaking load, kg per 1 sq. cm.

Distance from the ground (in feet).

Diagram 9.

Points 4 and 1, outlined with a dotted line, seem to indicate a deviation from the law of the decrease of strength toward the top; but this phenomenon should be regarded as accidental. It is possible that there was an incorrect recording of the machine readings here, or an error in the calculation.

An analogous phenomenon of a decrease in strength is observed beyond the elastic limit in bending of beams, but to a much more pronounced degree. Here, as can be seen from diagram No. 8, there is obtained an almost straight-line decrease in strength from 976.6 kilograms to 585 kilograms.

I shall not present diagrams of the modulus of elasticity, since from the tables it is easy to trace its decrease as the cut beams approach the top.

From diagram No. 9 the same dependence is also evident for axial compression (along the fibers). The same phenomenon is observed in tension as well. As for compression perpendicular to the direction of the fibers and shearing, here there is no law of decrease in strength in the direction toward the top of the tree. With respect to shearing one may still suppose that the cementing force of the fibers is the same throughout the entire trunk (shearing must be regarded as a separation of the fibers of the wood); but in compression perpendicular to the fibers it is quite impossible to find a suitable explanation for the absence of a law of decrease in strength. Without in any way claiming the infallibility of my investigations, I leave this question open.

III.

Influence of Moisture on Strength.

Bending.

For experiments of this kind one usually selects such a section of the trunk that the influence on strength in two adjacent logs, depending on the distance from the base of the tree, would not be manifested in the beams, or would be relatively negligible. The beams are taken in such a way that they are of identical structure, which implies the same density of the material. If the difference in age of two adjacent logs is comparatively large, then one confines oneself to eight beams from one log. Usually, however, up to 16 beams from two adjacent logs are tested.

The specimen tested is characterized by the following data:

Yellow pine (Pinus palustris) from Valas Province, State of Alabama.

Diameter . . . . . . . . . . . . . . . 18½ inches.
Height to the first branches . . . . 60 feet.

Age . . . . . . . . . . . . . . . . . . 190 years
Total height . . . . . . . . . . . . . . 110 ft.

The tree was sawn into 8 logs. For the test, two logs were taken at a height between 19 and 48 ft. The dimensions of the logs were: of the 1st: \(D = 18\) in., \(L = 13'4''\); of the 2nd: \(D = 17''\), \(L = 12'8''\). The sawn-out beams were subjected to artificial drying (except for beam No. 4 of the 2nd log), gradually lowering the percentage water content; then they were worked down to within \(3 \tfrac{1}{2}\) to 4 inches on a side and, after each beam had been sawn in two with the calculation that their length between points of support could be taken as 60 inches (1524 mm), the beams were sent for testing. With a decrease in the moisture of the wood from 44.2% to 11.5%, there occurred a corresponding decrease in specific gravity from 0.92 to 0.73 and an increase in strength, as is seen from Table 5. It must, however, be noted that the increase in strength in connection with the decrease in specific gravity is observed only for beams of identical structure; i.e., in the present case it may be said that strength increases with decreasing moisture. Generally speaking, greater specific gravity corresponds to greater strength at one and the same percentage water content. Below we shall see that, for one and the same species, sapwood is less strong than the heartwood portions, and now let us turn to the diagrams characterizing the bending of beams at different water contents.

Diagrams Nos. 10 and 11 show the fall in strength in bending within and beyond the elastic limit. In the present study, with a decrease in moisture from 44.2% to 11.5%, the resistance within the elastic limit increased by almost 80%; the ultimate load up to the moment of failure increased by more than 90%. And, finally, the modulus of elasticity increased (diagram No. 12) by 48%.

Compression in the direction of the fibers (axial).

For these experiments, from the same tree the second log from the base was taken, at a distance of 16–19 ft., with a diameter of \(18 \tfrac{1}{2}\) inches. The cutting of the beams was carried out in the same way as for bending, i.e., the log was sawn along mutually perpendicular diameters.

Dimensions of the tested beams: \(h = 210\) mm, \(a = 75\) mm, \(b = 75\) mm.
The percentage water content varied from 46.0 to 12.5%.

The results of the 21st test are summarized in the following table.

Compression.

Table 5-a.

Logs, Nos. and shape of cutting of the bars	No. of bar	Moisture, %	Specific gravity	\(l\), mm	\(h\), mm	\(b\), mm	Ultimate load in kg/cm, \(K'_b=\dfrac{P_1 l}{4W}\)	Elastic load in kg/cm, \(K_b=\dfrac{P_1 l}{4W}\)	Modulus of elasticity
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	1	11,5	0,73	1524	88,2	90,3	930,5	774,9	154500
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	1	13,4	0,76	1524	89,0	88,2	870,0	750,0	150000
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	2	17,1	0,77	1524	90,1	89,4	850,0	707,5	142500
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	2	18,4	0,77	1524	87,4	91,2	720,1	650,0	126000
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	3	19,0	0,78	1524	88,3	90,4	710,0	603,0	122500
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	3	19,2	0,78	1524	89,1	87,5	700,0	582,6	122000
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	4	21,9	0,78	1524	87,9	86,9	668,2	550,0	121000
[[diagram: circular log cross-section divided into four bars numbered 1–4]]	4	23,8	0,785	1524	89,2	90,0	650	540,2	119500
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	1	24,1	0,78	1524	86,9	88,2	635,2	538,0	119000
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	1	28,0	0,80	1524	89,0	87,6	634,0	521,4	117500
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	2	28,5	0,81	1524	88,6	89,4	620,0	513,5	116000
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	2	28,8	0,79	1524	87,8	87,4	610,4	503,1	113000
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	3	33,5	0,82	1524	89,3	87,9	560,0	463,5	115000
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	3	38,5	0,85	1524	90,2	91,2	552,1	462,8	108000
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	4	42,5	0,89	1524	86,8	88,4	525,1	432,1	106000
[[diagram: two longitudinal rectangular bars marked “D. 10” and “D. 11”]]	4	44,2	0,92	1524	91,3	88,9	480,2	425,0	105500

Diagram 12.

Vertical axis: Moisture, %.
Horizontal axis: Axis of moduli of elasticity.

Diagram No. 12, showing the dependence between the moisture content of the wood and the modulus of elasticity.

BENDING

Diagram 10.

Vertical axis: Elastic load per 1 sq. in.
Horizontal axis: Axis of moisture content, in %/o.

Diagram No. 10, showing the dependence between the moisture content of the wood and the elastic load.

Diagram 11.

Vertical axis: Modulus of rupture (per 1 sq. in.)
Horizontal axis: Axis of moisture content, in %/o.

Diagram No. 11, showing the dependence between the moisture content of the wood and the breaking force.

Table 5.

Moisture percentage	12.5	13.1	14.0	16.9	19.8	20.8	21.5	22.5	23.0	23.5	24.1	25.5	27.1	29.7	33.1	34.5	37.2	38.1	40.9	44.3	46.0
Breaking load per 1 sq. cm, kg	530.0	525.8	500.5	450.3	442.9	410.1	416.1	412.2	412.8	359.7	380.2	390.0	374.1	362.5	350.9	338.0	342.0	327.1	314.2	312.0	306.1

The law of the increase in strength as a function of the decrease in the percentage water content is sufficiently visible in Diagram No. 13, which expresses the table given above, No. 5. The increase in strength reaches more than 70% (for the present test).

Compression perpendicular to the direction of the fibers.

From the same tree a log was taken at a height of 5–16 feet from the base. The first part of the log, 7 ft long, was sawn into bars for tests in compression perpendicular to the direction of the fibers, and the second—5 ft long—for tests in shear.

The dimensions of the bars subjected to compression were: \(L = 410\) mm, \(a = 75\) mm, \(h = 100\) mm. The loaded area was \(75 \times 100\) mm. Number of tests—18.

Table 6.

Moisture percentage	12.5	15.0	16.2	17.0	18.6	20.55	21.25	23.8	25.6	27.8	29.4	31.25	35.8	36.8	38.7	40.9	46.1	48.1
Breaking load per 1 sq. cm, kg	90.4	88.2	78.8	70.2	74.1	66.5	68.2	61.9	55.0	52.0	47.8	46.2	43.8	42.3	37.4	34.1	33.2	31.0

Diagram No. 14 clearly expresses the dependence between moisture content and the resistance of the wood to compression perpendicular to the direction of the fibers. As is seen from the table, the strength increased by 196%.

Shear.

To determine the dependence between moisture content and resistance to shear, 22 tests were made. The tests were carried out by the method described above; the shear area was from 6.5 to 8.0 sq. cm.

Table 7.

Moisture percentage	12.4	14.3	15.3	15.5	16.5	17.0	17.5	19.0	19.1	19.2	23.5	23.8	26.5	27.0	28.9	30.9	32.8	34.0	37.3	39.8	44.5	48.5
Shearing force, kg per sq. cm	72.1	70.0	71.5	65.0	60.5	62.5	55.0	59.5	55.1	53.5	50.6	44.5	42.5	38.6	37.4	33.5	37.0	34.0	27.4	37.0	24.6	22.1

COMPRESSION

Diagram 13.

Vertical axis: Percentage of water.
Horizontal axis: Breaking load in kilograms per 1 sq. cm.

Diagram No. 13, showing the relationship between moisture content and the breaking load along the fibers.

Diagram 14.

Vertical axis: Water, %%.
Horizontal axis: Load axis in kilograms per 1 sq. cm.

Diagram No. 14, showing the relationship between moisture content and compression perpendicular to the direction of the fibers.

Thus, for compression, the strength increases as the moisture content decreases more than for the other kinds of resistance; and, for these 22 experiments, the increase in resistance reaches 227%.

Owing to lack of time, I was unable to carry out analogous experiments in tension, i.e., an investigation of tensile strength as a function of moisture content. But it must be supposed that the law of variation of strength will be the same.

Diagr. 15.

Diagram No. 15, showing the dependence between moisture content and the compressive condition.

IV.

Investigation of strength as a function of the distance from the axis of the trunk.

The pine specimen tested for determining strength as a function of the distance from the axis of the trunk was taken from Escambia Province, State of Georgia.

The tree is characterized by the following data:

Age	198 years.
Greatest diameter	30¼″ (stripped of bark
Total height	120 feet
Height to the first branches	60 feet.

For the tests, two logs were cut, each 12′8″ long, at a height from 10′ to 35′4″, with diameters of 29″ and 27.5″. I succeeded in completing the test only of the first log, and only in bending; therefore I confine myself to the bending table.

For these investigations the log was sawn into 64 beams measuring 64″ × 4″ × 4″ and 64″ × 3.5″ × 3.5″. Consequently, the cross-sectional area of the log yielded 32 beams, as shown in Figure 8.

After the sawn beams had dried, their moisture content ranged between 14.5% and 17.5%, to which corresponded a specific gravity of 0.58 to 0.66. And, finally, after planing on the planing machine

Timber No.	Breaking load	Ultimate elastic load	Modulus of elasticity
2	704,0	540,0	112000
3	662,8	582,4	109000
5	588,2	475,5	98500
6	702,5	563,0	108000
7	710,3	590,2	112300
7	870,1	702,0	130000
8	820,2	605,0	119200
9	772,0	623,0	113000
11	541,0	455,2	101000
12	822,2	670,0	131000
13	841,0	620,0	122600
14	954,0	880,0	160000
15	743,0	595,0	112900
16	729,3	615,0	119000
17	683,0	465,0	98500
18	705,8	492,4	112000
19	871,0	760,0	152000
20	865,4	702,8	132000
21	795,0	599,0	129000
22	704,7	570,0	118000
23	598,0	466,0	112800
24	790,0	632,0	125000
25	880,1	605,1	133000
26	751,0	625,0	118000
27	770,0	570,2	128000
28	702,0	591,5	106500
29	486,5	423,0	105400
29	795,0	529,4	106600
30	772,0	591,0	119000
31	700,5	608,0	110500
32	699,4	620,0	124000

Table 8.

the final dimensions of the timbers were: for the central ones—\(l=60''\) (between the points of support), \(h=3.5''—3.7''\), \(b=3.5''—3.7''\); for the outer ones—\(l=60''\), \(h=2.8''—3''\), \(b=2.8''—3''\).

Fig. 8.

The table 8 given here provides 31 tests, with a rejected timber from the first part of the log being replaced by the corresponding timber from the second part.

In preparing the timbers it was found that Nos. 1, 4, and 10 were unsuitable for testing (too small an area of the cross-section); for Nos. 7 and 29 repeated tests had to be made because of doubt as to the correctness of the reading.

In constructing diagrams of the dependence of strength on the distance of the timber from the axis of the trunk, we take the arithmetic mean of the data:

1) For timbers Nos. 13, 14, 19, and 20 (central).
2) ” ” Nos. 7, 8, 12, 15, 18, 21, 25, and 26.
3) ” ” Nos. 6, 9, 24, and 27.
4) ” ” Nos. 2, 3, 11, 16, 17, 22, 30, and 31.
5) ” ” Nos. (1, 4,) 5, (10), 23, 28, 29, and 32.

These arithmetic mean values for the breaking and elastic loads, as well as for the modulus of elasticity, are expressed in the table given below (Table 9).

Table 9.

	Breaking load.	Ultimate elastic load.	Modulus of elasticity.	Distance from the axis to the center of the scantling.
1)	878.0	740.0	141400	0.20 radius
2)	788.0	609.0	121900	0.44 radius
3)	758.0	597.0	118500	0.58 radius
4)	687.0	553.0	110800	0.70 radius
5)	644.0	517.0	108900	0.80 radius

From the data of this table we construct diagrams which clearly characterize the dependence of strength on the distance from the axis to the periphery of the trunk. The fifth column of the table gives the distances from the axis of the tree to the axes of the corresponding groups of scantlings; the calculations were made on the assumption that, for all scantlings, the sides of their transverse sections are equal to 4 inches, and that the radius of the log, as stated earlier, is 14.5 inches.

No. 16.

No. 17.

In No. 16 two diagrams are shown: AB, the diagram for the breaking load, and CD—for the ultimate elastic load. No. 17 gives the diagram for the moduli of elasticity. The numbers at the plotted points denote the number of scantlings, equally distant from the axis, that yielded the arithmetic mean in the final computation.

Such a sharp difference in the data on load and moduli (Table 9) gives us the right to say that the strength of the wood decreases as the bars are removed from the axis toward the periphery of the trunk. Of course, the number of experiments performed is not large enough to propose this as a generalization. But since we know that the heartwood portions of a tree are denser than the sapwood portions, we can already assert a priori that the central parts of a tree are stronger than the sapwood; and in this case the preceding experiments acquire the value of an argument in favor of our a priori consideration.

With this, strictly speaking, my investigation of the strength of yellow pine within the limits of the planned program comes to an end. The experiments are, of course, quantitatively few and in places do not cover all kinds of resistance (in the latter section). But the purpose of this article is to acquaint the Russian engineer with American methods of investigating the strength of wood and with the rational organization of the work of which the Washington laboratory may rightly boast. This laboratory has existed for only 12 years; nevertheless, almost all the typical representatives of construction in the United States have been examined in it.

One involuntarily turns to the activity of domestic mechanical laboratories for a comparative assessment of them and arrives at a sad conclusion. True, our engineers are far from the idea of making use of the directions of technical quacks concerning the properties of wood as a building material; but the data of the traditional Hütte—even if only from Bauschinger’s investigations—to which our ordinary engineer inevitably has to turn, can they really be applicable to materials of Russian local species, scattered under various climatic and soil conditions, such as, for example, Crimean pine (Pinus taurica) and Siberian pine (Pinus sylvestris)? Yet we know of laboratories which, in twenty and more years of existence, have not published a single scientific experimental investigation.

I would be fully satisfied if this small article aroused interest in the investigation—so necessary among us—of wood as a building material, and if Russian laboratories would open their doors to unglamorous but necessary work. It must be remembered that our country is the richest in the world in forest resources—300 million desyatinas!

P. A. Kozmin.

APPENDICES

to the article “Study of the Strength of Yellow Pine.”

DESCRIPTION OF THE MATERIAL.

Series No. 1.

Tree felled by the laboratory.

Series No. 2.

Tree purchased on the market
or obtained for testing by order.

1) Testing station: ____
Mechanical Laboratory of the Tomsk Technological Institute.

2) Timber number: 62—A—I.
Timber number in the order of testing in the laboratory — 540.

3) Common name: ______
Yellow pine, larch, etc.

4) Botanical name: ______
Pinus palustris, Pinus taeda.

5) Dimensions of the timber: ______

6) Acquired at ______
(Exact place of felling or of the market.)

7) Time of felling and location: ____________

8) Natural or artificial drying (with indication of time).

9) Comparative market valuation: ____________

10) Defects of the timber: *)
Side \(a\) _ (number of knots and cracks).
    ” \(b\) _ ” ” ” ”
    ” \(c\) _ ” ” ” ”
    ” \(d\) _ ” ” ” ”

11) Percentage amount of sapwood (by volume and approximately) ____

12) Rate of growth:

13) Photograph of the timber tested (Whether there is one or not. If there is, what kind).

14) When the form was filled out: ______

15) Remarks: __________

______ Signature of the person conducting the test: ____

*) A sketch of the timber must be on the reverse side of the form. For the sketch, two squares are made (usually by printed drawing) for the end faces of the timber, in order to show the place from which it was cut out of the log, and four rectangles across the full width of the form, on which the arrangement and number of knots are shown.

DETERMINATION OF MOISTURE CONTENT

1) Testing station: ................................ When tested: March 20,
1905.

2) Disks attached to beam No. ........ Species of wood:

3) Mark on disk: 62—B—1.

4) Nature of test: ........................ Bending, tension, etc.

5) When disks were cut:

6) When delivered to the laboratory ....................

7) When the moisture content was determined:

SKETCH OF DISK *)

Mark or disk No.	Thickness of disk	Weight of disk	Weight of disk	% of liquid bodies	% of liquid bodies
Mark or disk No.	Thickness of disk	Fresh-cut	Dry	Water	Resin or oil
62—B—1	1″—6″	125 kg.	64 kg.	43.5%
62—B—1	1″—6″	125 kg.	64 kg.	39.9
62—B—1	1″—6″	125 kg.	64 kg.	39.2
62—B—1	1″—6″	125 kg.	64 kg.	42.6
62—B—1	1″—6″	125 kg.	64 kg.	44.1
62—B—1	1″—6″	125 kg.	64 kg.	38.9
62—B—1	1″—6″	125 kg.	64 kg.	40.8
62—B—1	1″—6″	125 kg.	64 kg.	45.2

7) Average result for percentage of moisture content ............................

9) Remarks: ............................................................

............................................................

Signature of the person conducting the experiment: ............................................

*) On the sketch are shown the place and the distance from the center of the tested parts of the disk.

BENDING TEST.

Testing station ........................................
Specimen number: ........................................ Species: ........................
Dimensions of specimen: ................................. Full weight: ....................
Dimensions of specimen between points of support (span — \(l\)) ........
Machine:—(firm or construction and capacity .......................................

When tested: ........................
Laboratory test No.: ................
Moisture content in %: ........................
Specific gravity: ........................
Speed of machine per minute: ....................

POSITION IN THE MACHINE*).

(Sketch)

Here the end portions of the specimen are to be drawn by hand, with an indication of the arrangement and number of annual layers.

TEST RESULTS.

Ultimate elastic load: ........................ kilogrammes per 1 sq. cm.
Breaking load: ................................ kilogrammes per 1 sq. cm.
Shearing forces: .............................. kilogrammes per 1 sq. cm.
Modulus of elasticity: ........................ kilogrammes per 1 sq. cm.
Total work of loading: ........................ kilogramme-metres.
Work of loading up to the elastic limit: ...... kilogramme-metres.

Work per cubic centimetre or decimetre (inch) of the specimen:

Elastic load: ........................	kilogramme-metres.
Ultimate load: .......................	kilogramme-metres.

Duration of test.	Load.	Mirror or scale readings.	Deflection arrow	Remarks

*) The sketch of the fracture is shown on the reverse side of the blank, where for this purpose four rectangles are printed across the full width of the page. There also are made the calculations of loads, modulus, etc.

Experimenter ........................................

Head of laboratory ........................................

COMPRESSION TEST.

Compression along or across the fibers.

Testing station: ........................................................
Specimen No. ................................................
Mark on the specimen tested: ........................................
Height: ............ Cross-sectional area: ................................
Machine: ................................................................

When tested: ........................
Species: ........................
Laboratory test No.: ........................
Weight: ............ Moisture content: ............
Speed of the machine: ........................

SKETCH.

The end faces of the specimen are shown.

TEST RESULTS.

Elastic-limit load: ................................................ lb. per 1 sq. in.
Breaking load: ...................................................... lb. per 1 sq. in.
Modulus of elasticity: .............................................. lb. per 1 sq. in.
Specific gravity: ....................................................

Duration of the test.	Load.	Reading.	Amount of compression.	Remarks.

SKETCH OF THE SPECIMEN AFTER THE TEST.

Here all four side faces are shown, with the effects of failure.

Experimenter ........................................

Head of the laboratory ........................................

SHEAR TEST

Testing station ....................................................
From timber No. ....................................................
Mark on the timber being tested ....................................
Shearing area: .....................................................
Number of annual rings per inch: ...................................

When the experiment was performed: ................................
Species: ..........................................................
Laboratory test No. ................................................
Weight of timber: ............ Moisture content ....................
Machine: ...........................................................

SKETCH.

TEST RESULTS.

Shearing stress for
{ radial shear ........................................ kg per 1 sq. in.
{ tangential shear: ................................... kg per 1 sq. in.

Duration of experiment.	Load.	Remarks.

SKETCH OF THE TIMBER AFTER THE EXPERIMENT.

Experimenter: .....................................................

Head of laboratory: ...............................................

Mechanical Laboratory

at Washington University in St. Louis, Mo.

The area of the laboratory is \(71' \times 46'\). The machines for working the timbers prepared for testing are driven by a steam engine and an electric motor by means of the main shaft \(A\). Of the woodworking machines, the laboratory has: two planing machines (\(a\)—planing on four sides, \(b\)—single-drum), \(c\)—a mortising-and-drilling machine, \(d\)—a lathe, \(e\)—a circular saw, \(f\)—a band saw.

Testing machines: a Riehlé universal machine for 50,000 kg; a machine for bending small timbers for 50,000 kg; a machine for bending large timbers for 100,000 kg; a machine for testing cement; a machine for testing large timbers in compression.

In addition, the laboratory contains a drying room heated by exhaust steam.

Riehle universal machine for 50,000 kilograms.

Lever B—for starting the machine; D and C—for speeds within the range from 2.54 mm/m (minimum) to 50.8 mm/m (maximum) per minute; A and D—levers for setting speeds up to 203.2 mm/m per minute.

Olsen machine for 100,000 kilograms, installed in the Mechanical Laboratory of the Tomsk Technological Institute.
For a detailed description, see the drawing.

Fig. 1.

Fig. 2.

b—main column for tensile tests; a—guide column; e—automatic diagram apparatus; c—device for rotating the diagram drum; d and d₁—electrical contacts for the forward and reverse motion of the load; h—handwheel for the lever with friction drive, for changing the speed of the machine.

Olsen testing machine for 100,000 kilograms.
Washington Laboratory and the mechanism of the laboratory of the Tomsk Technological Institute.

\(F\)—platform transmitting the pressure to levers \(F_1\) and \(F_2\);
\(F_2\)—outer lever.

\(h\) and \(h_1\)—platform knife-edges for levers \(F_1\) and \(F_2\).

\(e\) and \(e_1\)—knife-edges for levers \(F_1\) and \(F_2\), resting on the frame of the machine.

\(g\) and \(g_1\)—knife-edges of levers \(F_1\) and \(F_2\).

\(f\) and \(f_1\)—knife-edges of the middle lever \(F\).

The ends of levers \(F_1\) and \(F_2\) rest on the frame of the machine.

Lever arms:

\(gh = 114.5\ \mathrm{mm}\); \(eg = 1834.5\ \mathrm{mm}\)—for the outer levers.

\(fe = 992\ \mathrm{mm}\); \(f\varphi = 1477\ \mathrm{mm}\)—for the middle lever.

\(F_3\)—the second lever, resting on prism \(K\); \(F_3\) is connected by rods \(l_2\) and \(F_2\) to levers \(F_1\) and \(F_2\) with the help of a straightedge \(m\), the ratio of which is \(l_3 = 77\ \mathrm{mm}\), \(l_m = 955\ \mathrm{mm}\).
Shoulder of lever \(F_3\): \(kl = 77\ \mathrm{mm}\), \(km = 955\ \mathrm{mm}\).

On the suspended lever there are:

Load \(G\), which receives automatic displacement by means of the rotation of the screw on which the load is mounted with the aid of a nut. The speed of motion of the load is regulated by conical pulleys.

Load \(b\)—for balancing the weight lever.

Load \(d\)—for balancing the self-weight of the beam and for setting the machine in the unloaded state.

\(B\) serves for starting the machine in forward and reverse motion.

\(Q\)—for changing speeds; \(T\) and \(T\) are the pulleys for receiving motion from the transmission.

The movable table serves for reverse motion.

The dimensions of the shoulder of the plate are established in the mechanism in the workshop laboratory of the Tomsk Technological Institute. The other details of the machine according to the photograph have the same appearance as that of the machine adapted for bending tests.

1. Mittheilungen aus dem Mechanisch-Technischen Laboratorium, Mittheilung XIX, München, 1887. ↩

2. A supposition expressed by Profs. N. M. Kizhner and V. N. Dzons. Tomsk Technological Institute. ↩