Two sets of data are analyzed in the thesis. The first set was derived from the long term deformations of 2 in x 6 in x 12 ft (40 mm x 140 mm x 3600 mm) joists of Douglas-Fir loaded under constant bending stress to levels lower than or equal to 3110 psi (21.44 MPa). The second set was derived from the long term deformations of 2 in x 6 in x 12 ft joists of Hemlock loaded under constant bending stress to levels of 3000 psi (20.68 MPa) and 4500 psi (31.0 2 MPa) . The analysis shows that the creep behaviour of structural size beams depends upon the material characteristics; specifically, material with a strength lower than 5000 psi (34.33 MPa) appeared to creep 1.5 times more than, material with a strength higher than that level, over a three month period. In addition, the test results support the assumption of a linear relationship between the creep deformation of a structural-size timber beam and applied stress. A method is presented to predict the creep behaviour of a structural-size specimen at discrete times over a three month period. The method consists of expressing the creep deformation, Δ[sub c], in terms of the elastic deformation, δ[sub e], or equivalently, the fractional creep (f = δ[sub c] / δ[sub e]) in terms of δ[sub e] or in terms of the modulus of elasticity. This work is limited to the stress levels investigated and to specific temperature (10°C< θ <30°C) and moisture content (8%<MC<12%) conditions. While this method could be employed in preliminary design procedures, it has been especially designed for more complex studies of the creep behaviour of structures including floor systems, trusses, etc... The advantage of the method is that in this kind, of analysis the modulus of elasticity of the individual components can be used. This thesis also presents a set of creep curves that cover a three year span. These creep curves show that the average total deformation of beams loaded to a stress level of 3110 psi (21.44 MPa), at this time, is approximately 1.6 times the elastic deformation. / Applied Science, Faculty of / Civil Engineering, Department of / Unknown

To date there are no long-term creep models or practical methods to investigate the effect of creep on the safety and serviceability of modem wood structures and structural wood composites. Long-term creep models were developed for wood in tension and compression using the Time-Temperature Superposition Principle (TTSP). The principle states that the long-term response of a polymer at lower temperature is equivalent to the short-term response at a higher temperature. Accelerated creep tests were conducted in tension and compression using small clear specimens of Douglas-fir, southern pine and yellow-poplar. The specimens were tested at moisture contents of 6 %, 9 %, and 12 %, and at temperatures between 20°C and 80°C. The strain was measured using bonded strain gages. The individual creep compliance for each temperature was shifted along the log-time axis to obtain a "master" curve that describes the creep response of the specimens. All compliance curves also required vertical shifting. The experimental horizontal shift factors followed the Arrhenius formulation that describes the shift factor relation for polymers in the glassy region. / Master of Science