A timber specimens highest strength grade for a selected strength grade combination can be
determined by optimal grading. This can be achieved by different approaches that are
represented in the current standard and pre-standard. The pre-standard defines a parametric
and non-parametric calculation of characteristic properties from test results, which is the
essential difference between various approaches. In both the non-parametric approach and
the standard procedure, some simplifications and adjustment factors are used while in the
parametric calculation only equations from a theoretical background are used. Procedures are
derived for eight different strength grade combination and for all results comparisons are made.
With the Mathematica software we used simulations to test given equations for the calculation
of characteristic properties. Special consideration was taken on the non-parametrical
calculation. The correction factor equation has not proved to be suitable, at least not for the
assumptions that are made in the pre-standard. Therefore, we produced a new equation, which
gives better results.
A new strength grading machine is under development, which is not yet standardized and
cannot formally be used for strength grading and CE marking yet. The requirements for the
derivation of machine settings for initial type testing were met and we established new settings.
We tried different properties as the indicative property. Due to its good correlation to specimens
strength properties, dynamic modulus of elasticity was chosen. Grading procedure could be
simplified by avoiding density measuring. Using size matrix and global cost matrix, we
evaluated grading machine performance for data with measured density and data with
constant density for each of the strength grade combination. The results using the measured
density were significantly better than those with a constant density. The strength grading
machine gave good results and is therefore suitable for the certification process.
|
