Sample size effect on the structural properties of glue-laminated posts
1992
Wright, B.W. | Manbeck, H.B. | Troxell, D.F.
An investigation was conducted on the effect of sample size upon the calculated allowable structural properties of glue-laminated splice joints in wood construction posts. Based on the three-parameter Weibull distributions determined for 16 structural property data sets, cumulative distribution functions were used to randomly generate samples with sizes of 20, 40, 66, and 80. The generation of sample groups was repeated 1,000 times for each of the four sample sizes in each of the 16 data sets. Upper and lower bounds were determined for estimation of the structural properties. Factors were determined to conservatively adjust the structural properties for sample sizes less than 80. Adjustment factors ranged from 1.080 for a sample size of 20 to 1.020 for a sample size of 60 in the determination of allowable bending stress. Similarly, adjustment factors for sample sizes ranging from 20 to 60 varied from 1.083 to 1.044 and 1.100 to 1.012 for compressive stress and upper bound elastic modulus determination, respectively. The adjustment factors for lower bound elastic modulus determination ranged from 0.938 to 0.989 for sample sizes of 20 to 60, respectively. Factors of this type could provide a valuable and conservative alternative to testing a complete set of 80 samples for the determination of structural properties in laminated wood products.
显示更多 [+] 显示较少 [-]