This paper represents a new approach to size effects in timber. The design process in Canadian Standards Association code CAN3-086-M80 allows for size effects in shear and in tension but not in bending. A large number of bending test results arc analyzed in this paper to show that size effects in bending are very important. A modification is made to the traditional weakest link theory for size effects to reflect the anisotropic nature of timber. Good agreement is obtained between tests and theory. The theory enables us to compare tests with different spans and (or) load configurations with an understanding not previously possible. Results from several large testing programs are used to quantify the theory. For bending members, length effects and load configuration effects are found to be much more important than depth effects, and a simple design method is proposed. Information for tension members is less comprehensive; nevertheless, a tentative suggestion for the design of those members is included.

The paper points out that the present method used to dcsign bcaring platcs for wood subjcctcd to stresses pcrpcntlicular to the grain is inadequate because it ncglccts the cffect of the spccific gconletry. A design mcthod that more correctly rcflccts the actu;~l conditions is prcscnted together with test data that rcprescnt thc matcrial propcrtics in a more realistic way.Thc design method is discussed and comparcd with limit states design and dcsign examples arc presented. Thc mcthod includcs a conservative estimatc of thc dctlcctions.Can. J. Civ. Eng., 9, 338-349 (1982)

This paper deals with some of the problems arising when the strength of a structural member varies along its length such as occurs in timber members where knots and other natural growth characteristics create cross sections with varying strengths along the length of the member. Failure may then take place wherever a weak cross section happens to be subjected to high stresses. A different concept of strength must be introduced into the design codes in order to reflect the probability of such an occurrence. Strength becomes a function of length, depth, and load configuration, all of which must be considered in the design process if both economical and safe structures are to emerge. In this paper, which was directed toward establishing the length effects for 38 mm thick spruce–pine–fir in tension and compression as well as bending, the results of a testing program are described. It was, for instance, found that when the length of tension members is doubled from 3.0 to 6.0 m, the strength reduces to 86% of the previous strength, or if the length is quadrupled (as may occur in the bottom chords of trusses), the strength could reduce to 73% of the original strength. A review of the available research information is presented and evaluated in order to assess the adequacy of the available size effect information for design purposes. The present size effect requirements incorporated in some existing timber codes are also reviewed. Key words: timber, strength, size effects, length effect, weakest link theory, design method, load configuration, bending, compression, tension.

This paper describes experimental work on size effects occurring in defect-free Douglas fir and suggests empirical methods for modelling these. It was found, contrary to present belief, that the size effects could best be described as a volume effect. It was also found that wet and dry materials behaved very differently with respect to size effects. Defect-free material represents the stronger and more expensive range of materials available to the structural engineers. It is, therefore, important that the material behavior be better understood in order to create structures that are both safe and economical. Key words: size effects, clear wood, defect-free wood, volume effect, length effects, load configuration effect, moisture content, weakest link theory.

The present allowable stresses for lumber and the associated adjustment factors stated in Canadian Standard CSA-086 are based upon tests conducted using small clear wood specimens. Recent comprehensive tests in which full-size lumber specimens were used (in-grade testing) showed that the present allowable stresses cannot be confirmed and also that new adjustment factors need to be developed.This paper deals with the effect of moisture content on the strength properties. Tests were conducted and reported dealing with: bending, stiffness, tension, compression parallel to grain, compression perpendicular to grain, and shear. From these tests it was possible to make specific recommendations for a new set of moisture content adjustment factors, to be used in design, that more realistically reflect the behaviour of lumber.