EN 338 Strength Grade and Uncertainty Models of Material Properties for Nigerian Grown Terminalia Superba (White Afara) Timber Specie

  • JM Kaura
  • I Abubakar
  • I Aliyu
  • u Hassan
Keywords: Bending strength, Coefficient of variation, Density, Modulus of Elasticity, Strength class, Terminalia superba, Uncertainty


In this paper, the EN 338 strength class of Nigerian grown Terminalia superba timber specie was determined. This will enable the utilisation of the specie in the design of timber structures using Eurocode 5. The uncertainty models (mean, coefficients of variation and theoretical distribution models) of the tested specie were also generated. This information is required if structural reliability analysis and design is to be undertaken. The strength classification and the material properties were based on detailed laboratory experiment conducted on samples of the tested timber specie based on EN 408 (2004) and EN 13153-1 (2002). The laboratory data was analysed using an open source statistical computer package, Easyfit (2010) in accordance with EN 384 (2004). Terminalia superba was found to be of strength class D24 according to EN 338 (2009). The statistical analysis revealed that the coefficient of variation of modulus of elasticity is 21%. The coefficient of variations of both the bending strength and density is 19%. It was also found that, the best theoretical distribution models for density is normal distribution, while lognormal distribution was found to be the best distribution model for both bending strength and modulus of elasticity.  Due to the high coefficients of variation of material properties of timber as indicated in this study, it is recommended that EN 338 strength class should include the information on the statistical parameters of timber. With this, the design criteria of timber structures using Eurocode 5 can be enhanced by the use of structural reliability method in order to fully accommodate uncertainties.


Building, Civil & Geotechical Engineering

Journal Identifiers

eISSN: 2467-8821
print ISSN: 0331-8443