STRENGTH OF BUILT-UP TIMBER COLUMNS

Abstract

At the present only a limited amount of information is provided by various timber codes and specifications for the design of built-up timber columns. Very little information is provided for designing layered timber columns and no information is given for braced columns. The design method provided for spaced columns is based on empirical tests and formulas. The objective of the present investigation is to develop a rational procedure for the analysis and design of mechanically connected built-up timber columns including layered, braced and spaced columns. The theory developed takes into account the effect of interlayer slip and is applicable to columns failing in the elastic as well as inelastic ranges of stress. The test material is Construction Grade No. 1 .Eastern Spruce lumber. Connector types used are common wire nails, steel bolts and split ring connectors. The philosophy and the procedure presented herein are applicable to other species of wood as well. The investigation is conducted in three phases. The primary aim of phase one is a general theory for predicting the load-slip behaviour of timber joints subjected to interlayer slip. Some 250 connections fabricated from three to seven members and fastened with various types of connectors are tested. Good agreement is observed between the theory and the experimental results; the overall average difference between the two is about five percent. In addition, 2,130 compression tests are conducted to evaluate physical, strength and elastic properties of the test materials. Phase two provides a general theory, incorporating the results of phase one, for predicting the buckling stresses of built-up timber columns. To verify the theory, a comprehensive test program is conducted on some 400 columns, including layered, braced and spaced columns of various dimensions built up from two to seven members and covering the range of slenderness ratio values from thirty-four to one hundred and fifty. The cross-sections investigated have one or two axes of symmetry. Statistical techniques are applied to analyse the test data. Good agreement is observed between the theoretical predictions and the experimental results. The overall average difference observed between the predictions and the experimental results is about six and one-half percent. Incorporating the results of phases one and two in phase three, a rational procedure, using dimensionless coefficients called 'Buckling Coefficients', is developed for the de s ign of built-up timber columns. This design method is simple to apply and is applicable to elastic as well as inelastic columns.