Woodflanges under Tension

The tensile strength of wood in small dimensions has been investigated experimentally. Tests have been carried out partly with regular tension test specimens, and partly with glued beams with thin webs and where the local lateral deflection of the flanges have been controlled more or less. This control originated in partly an effect from the beam itself, partly bracing from glued-on panels, and it results in an decrease of the maximum stress and strain around knots, and so causing an increase in the apparent tensile strength.

The results show an increase in the mean tensile strength of 20 to 40 percent when determined by bending tests. The largest strength increases occured from beams where the local deflections of the flanges were controlled mostly. Further, it was found that the standard deviations in beam tests were less than those found in regular tension tests. Both effects lead to that the characteristic tension strengths found in bending tests are much larger than those found in regular tension tests. The increase is 30 to 70 percent for flanges of the better quality T24 and 50 to 80 percent for the lower grade u/k.

Further, it was found that one gets almost the same tension strength of a flange in an I-beam whether it is made of two pieces produced by split-sawing one graded batten into two.