TIMBER PHYSICS FACTOR OP SAFETY. 



365 



The directions for using these indeterminate factors of safety given in the text-hooks would imply that the 

 student or engineer is, after all, to rely on his judgment as to the modification of the factor, i. e., he is to add to this 

 general guess his own particular guess. The factor of safety is in the main an expression of ignorance or lack of 

 confidence in the reliability of values of strength, upon which the designing proceeds, together with an absence of 

 data upon which to inspect the material. With a larger number of well-conducted tests, coupled with a knowledge 

 of the quantitathe as well as qualitathe influences of various factois upon strength, and with definite data of 

 inspection which allow ready sorting of material, the factor of safety , as far as it denotes the residuum of ignorance 

 which may he assumed to remain, as to the character and behavior of the material, may be reduced to a minimum, 

 restricting itself mainly to the consideration of the indeterminable variation in the actual and legitimate application 

 of load. 



Besaltb of ttsts in compression auhiise on green wood (above 40 per cent moisture, not reduced), 



[Pounds per square inch.] 



No 



1 



2 



3 



4 



7 



S 



9 



11 



12 



14 



16 



19 



20 



21 



22 



23 



25 



26 



27 



32 



Spec ies. 



Longleaf pine 



Cuban pme 



Shortleaf pine 



Loblolly pine 



Spruce pine 



Bald cypress 



White cedar 



"Wliiteoak 



Overcup oak 



Cow oak 



Texan oak 



"Willow oak 



Spanish oak 



Shagbark hickory . . 

 Mockernut hickory 



Water hickory 



Nutmeg hickory . . . 



Pecan hickory 



Pignut hickory — 

 Sweet gum 



Number 

 of tests. 



Highest 



single 



test. 



86 



38 



8 



69 



71 



280 



34 



25 



45 



58 



39 



49 



52 



22 



18 



4 



26 



4 



5 



6 



7,300 

 6,100 



4, 000 

 5,500 

 4,700 

 8,200 

 3,400 

 7,000 

 4,900 

 4,900 

 6,000 

 5,500 

 5,100 

 6,900 

 7,200 



5, 600 

 5,500 

 3,800 

 6,200 

 3,600 



Lowest 



single 



test. 



2,800 

 3,500 

 3,000 

 2,600 

 2,800 

 1,800 

 2,300 

 3,200 

 2,800 

 2,300 

 3,100 

 2,300 

 2,500 

 3,500 

 4,500 

 4,700 

 3,700 

 3,300 

 4,700 

 3,000 



Average 

 of all 



tests. 



4, 300 

 4,800 

 3,300 

 4,100 

 3,900 

 4,200 

 2,900 

 5,300 

 3,800 

 3,800 

 5,200 

 3.800 

 3, 900 

 5,700 

 6,100 

 5,200 

 4,500 

 3,600 

 5,400 

 3,300 



While the values given in these tables may claim to contain more elements of reliability than most of those 

 published hitherto, much more work will have to be done before the above-stated aim will be satisfied. 



In explanation of the table recording tests in bending at relative clastic limits it should be stated that since 

 an elastic limit in the sense in which the term is used for metals, namely, as a point at which distortion becomes 

 disproportionate to load and a permanent injury and set results, can not be readily determined for wood, Prof. J. B. 

 Johnson has proposed to utilize a point wheie the rate of distortion becomes 50 per cent greater for the amount of 

 load than it was for the initial load, which point can be tolerably accurately determined (see Bull. 8, p, 9). This 

 point he has called the "relative elastic limit." The assumption is that such a point would be near the limit to 

 which the material can be strained without permanent injury, and the strength values obtained at that point would 

 serve for indications of safe loads. 



The practical utility of determining this point and the strength values relating to it remains, however, still 

 open for discussion. A comparison of the values obtained for the strength at rupture and at relative elastic limit 

 shows a parallelism which would make it questionable whether much is gained by the use of that point, which in 

 reality lies beyond the limit where practical injury has begun, as indicated by the increased distortion. 



We would be inclined to consider that point more serviceable where the curve begins to deviate from the straight 

 line, at which point we may assume no permanent injury has as yet been experienced. This point we may call 

 provisionally the "safe limit." 



Objection has been made to utilizing this point because it can not bo located with as much nicety and mathe- 

 matical precision as the point of "relative elastic limit." But even this point is only approximately definable; and 

 since no strength values can claim to be more than approximately correct, it would suffice to determine the safe- 

 limit point and the correspondent strength values also only approximately. This point has the advantage that it 

 lies on the safe side. 



Special series of tests to investigate the legitimacy of the use of any of these limits for practical purposes 

 were designed, but have as yet not been taken up, and hence the values in the table on p. 367 are given only as 

 suggestions for what they are worth. 



