Experimental and numerical analyses are presented concerning of compression tests parallel and perpendicular to the grain, three-point bending, and double-shear joints in compliance with the relevant test standards (ASTM D2395, BS 373, EN 383 and EN 26891). Woods of Norway spruce (Picea abies L. Karst.) and European beech (Fagus sylvatica L.) were tested to describe their non-linear behavior. Elasto-plastic material models were the basis for the finite-element (FE) analyses with the input of own experimental data and those of the literature. The elasto-plastic material model with non-linear isotropic hardening was applied based on the Hill yield criterion in regions of uniaxial compression. The material characteristics were first optimized and validated by means of basic 3D FE models under the same conditions as applied for the experiments. Afterwards, the validated material models were implemented into the solver with more complex numerical analyses of wooden dowel joints. Concurrently, the digital image correlation (DIC) served for verification of the numerical wooden joint models. A good agreement (with a relative error up to 16%) was found between numerically predicted and experimentally measured data. The differences may be mainly attributed to some natural characteristics of wood which were not considered in the proposed material models. The proposed elasto-plastic material models are capable of predicting the wood’s ultimate strength, and therefore could contribute to a more reliable design of wood structures and their performance.

No abstract

The goal of the study is to investigate the non-standard deformation behaviour of wood loaded by compression parallel to the grain. This is represented as a negative increment of strain in the range of plastic deformations when the load continues to increase. The objectives of this study are to point out this problem and to provide its description based on the deformation fields that have been analysed using three approaches: a) full-field optical technique based on digital image correlation (DIC); b) “clip on” extensometer and its virtual analogy, and c) crosshead displacement method. Further, the negative strain phenomenon was studied depending on the sample length. The samples were made from the European beech (Fagus sylvatica L.) and Norway spruce (Picea abies L. Karst.). Based on the strain analysis, it can be concluded that the deformation field consists of three sub-regions exhibiting different stiffness values (three-spring model). The failure of less stiff zones near the compression plates during the “non-standard” compression behaviour causes almost zero compression deformation of the stiffer middle zone or even leads to its expansion. The three-zone heterogeneity of deformation field induces a deviation of the displacement and strain measured by the proposed approaches. This phenomenon substantially influences the resulting longitudinal Young’s modulus and, therefore, should be of concern when measuring wood in such mode.

The goal of the work was to evaluate mechanical performance of full-scale timber beams containing scarf joint with a dowel. Work focused on standard testing using modular system to obtain eff ective stiff ness and strength of the beams with and without the joint. The work further researched a contact zone between two timber parts of the joint -at the scarf face. This was carried out using non-destructive optical technique -digital image correlation (DIC) and newly developed algorithm. The joint was made of Norway spruce, dims. 6×0.2×0.24 m and was loaded by two modes: a) 3-point bending and b) 4-point bending. During the loading, a sequence of images was acquired for further investigation of contact zone using the proposed algorithm. The joint with scarf and dowel provided enough eff ective stiff ness, ie. 73-93% for 3-point bending test and 71% for 4-point bending with respect to MOE measured on reference solid beams. Eff ective strength of the joint was also relatively high and in a range of 55% and 60% with respect to reference solid beams in both 3-point and 4-point bending tests. Contact length diff ered for loading modes. Mean contact length in symmetrical 4-point bending was about 40%, for asymmetrical 3-point bending test, it was approx. 20% on face closer to support and 44% on a face closer to loads.

A sophisticated approach for the precise determination of both longitudinal shear moduli of wood at single test is introduced. The method is based on the combination of the torsion test inducing pure shear stresses in sample and an optical method providing the full-field strain data of such stress state. The proposed procedure of the longitudinal shear moduli determination consists of two main steps. In the first step, the apparent longitudinal shear modulus following the standardized procedure (EN 408?A1) was determined. Secondly, both longitudinal shear moduli were derived based on the apparent longitudinal shear modulus and the shear strain distribution on the radial and tangential sample surfaces. The wood of European beech (Fagus sylvatica L.) was used as material for the experiments. The exploratory analysis revealed the increasing difference between the longitudinal shear moduli determined in the longitudinal-radial plane and in the longitudinaltangential plane as the total torsion angle increased as well as with the increase in the average torsion stiffness. Further, the longitudinal shear moduli and the torsional longitudinal shear strength did not correlate well. Therefore, they cannot be used in order to predict each other. Although such findings need more detailed studies, they should be taken into account when designing wood structures.