Local variation in bending stiffness in structural timber of Norway spruce : for the purpose of strength grading

Sammanfattning: Most strength grading machines on the European market use an averagemodulus of elasticity (MOE), estimated on a relatively large distance along awood member, as the indicating property (IP) to bending strength. Theaccuracy of such grading machines in terms of coefficient of determination israther low at R2 ≈ 0.5. This research is motivated by a desire to increase theaccuracy of the strength grading in the industry today. The aim of the presentstudy is to contribute knowledge of local variation in bending stiffness/MOEwith high resolution and thus locate weak sections due to stiffness reducingfeatures (the most important is knots) for structural timber.The present study introduces three methods that involve structural dynamics,classical beam theory and optical measurement to assess local wood stiffness.Specifically:The dynamic method, in which a wood member is treated as an ordinaryphysical structure and the local stiffness is studied by exploring itsdynamic properties.In Method II, a bending MOE profile is established based on local fibre angle information. The local fibre orientation is detected through highresolution laser scanning based on the tracheid effect. For Method III, a bending MOE profile is established using surfacestrain information under four-point bending. A high resolution strainfield is obtained using the digital image correlation (DIC) technique.From the present study, the two latter methods are more favourable inevaluating the local stiffness within a piece of structural timber. Moreover, thestudy reveals that the established bending MOE profiles using the two lattermethods, i.e. based on information of the local fibre angle and surface strain,agree reasonably well. However, for some patterns of knot clusters, the localbending MOE, calculated on the basis of fibre angles, is significantly higherthan the local bending MOE estimated on the basis of surface strain.