Characterizing Reduction of Variability in Bamboo Axial Members Through Use of Multiple Culms and Grading

Bamboo culms possess considerable natural variability in geometry and material properties. There is a need to control this variability to achieve a reliable performance of bamboo structures. This study examines two methods for managing the variability: (a) the use of multiple bamboo culms and (b) grading. To that end, it first performs single-culm compressive strength tests on Kao Jue bamboo specimens. Then, it fits a nonlinear curve onto the experimental results using brittle, strain-softening, and elastic–plastic models and generates a large synthetic dataset by fitting a joint probability distribution model to the single-culm results. Through Monte Carlo simulations, the study uses the synthetic datasets of single-culm members to simulate the response of multi-culm members under compression loads. It then evaluates the statistics of the simulated multi-culm members. Findings reveal decrease in the coefficient of variability of capacity, stiffness, compressive strength, and modulus of elasticity properties follows an inverse square root function of the number of culms in multi-culm members for strain-softening, and elastic–plastic models. To evaluate the effect of grading, the study divides the same synthetic dataset of single-culm members into two sub-grades based on external diameter or stiffness. The statistics of the subgrades show that grading alters the distribution and reduces the variability of only those properties that are well correlated with the indicating property (i.e., the external diameter or stiffness). Overall, the two considered methods are effective in managing the natural variability of bamboo culms.