Modelling Young’s modulus on small clear specimens in relation to silvicultural treatment.

The aim of this research work performed in Task 11 "Modelling Young's modulus on small clear specimens in relation to silvicultural treatments" was the modelling of Young's modulus (MOE) on small clear specimens of 24 Belgian Norway spruces in relation with the silvicultural treatment. The aim of the research was also to compare Young's modulus on small clear specimens with Young's modulus on commercial size specimens with a view to defining the rate of prediction of wood quality of specimens including the natural timber defects.
This work is corresponding to one task among the twelve of a large ECC Project entitled "SilvicuItural control and non destructive assessment of timber quality in plantation grown Spruces and Douglas fir" for which 210 trees have been sampled through 6 different European countries.
The experimental raw material comes from the southern and eastern part<; of Belgium and concerns four different kinds of stands characterized by the site productivity class and the thinning intensity.
Within each stand, the trees have been sampled in three different social positions classes.
The small clear specimens have been marked off with reference to the radial and axial positions inside the 24 trees.
On the whole 350 clear wood specimens have been selected and tested by two different laboratories in order to get young's modulus by a non destructive four points flexural method described by the French Standard NF B 51-016.
Young's modulus appears to be positively related with specific gravity and negatively related with growth ring width. Stands growing on high productive sites or those submitted to low thinnings and also suppressed trees give rise to a stiffer wood. Among all these factors, only the social position of the tree within the stand influences significantly the Young's modulus values.
Nevertheless, a slow growth producing a heavier timber is not always linked with a stiffer wood.
Young's modulus differences between juvenile and adult wood vary in connection with the specific gravity class considered. In other words, juvenile wood is stiffer than adult wood when referring to low density classes, but adult wood becomes stiffer than juvenile wood when referring to high density classes. That means that the stiffness of adult wood is more sensitive to changes in specific gravity than the stiffness of juvenile wood.
The distribution of Young's modulus in relation with specific gravity shows a higher variability when specific gravity increases. In searching the reasons for a higher Young's modulus variability when specific gravity increases by looking back to the population of clear specimens, we point out a very low slope of grain on some specimens and above all the presence of compression wood on other specimens.
When rejecting these defects, then it appears that Young's modulus is more strongly related to specific gravity, thus giving rise to more accurate models.
Among all these models, the best one is corresponding to a combination of variables such as specific gravity and growth ring width per sampling level within the tree.
The variance analysis of residues of each model has revealed a tree effect always very highly significant. On the opposite, the other factors, as site productivity, thinning intensity and social position of the tree within the stand, have no significant effect on Young's modulus values.
That means that inter trees variability is far higher than the other sources of variation (intra tree variability, site productivity variability, thinning intensity variability and social position variability).
This also means that the general model established is valid whatever the silvicultural treatment could be.