Hardwood in European mixed forests will increase in response to climate shifts and thus investigating their properties gains prominence in research. Wood, whether softwood or hardwood, is hierarchically organized and the macroscopic, anisotropic mechanical properties are governed by the microstructure. The presented application of multiscale models aims to explore the influence of microstructural characteristics for an explanation of observed macroscopic variations in experimental testing of stiffness and strength perpendicular to the grain and rolling shear. A validated multiscale micromechanical model for the stiffness of softwood was presented by Bader et al. (2011) and further extended to hardwood by de Borst and Bader (2014). In the latter study, the model was validated for beech and nine other hardwood species. In this work, the mechanical properties of the softwood species spruce as a reference and the hardwood species beech and birch, are investigated. The experiments were conducted on prismatic clear wood specimens (50 mm width, 20 mm thickness (grain direction), and 60 mm height), enabling clear wood material-level testing. However, due to thickness and annual ring distribution of the raw material, the inclination of the annual rings of early- and latewood, defining the radial and tangential direction (RT-plane) of the idealized orthotropic material, varies in the test series. The influence of this RT-plane rotation and the variation in microstructural parameters, like microfibril angle, lumen, vessel, and ray cell properties, on the global stiffness, are modelled based on multiscale models for the three investigated species. The micromechanical models are applied to obtain values of the nine independent components of the stiffness tensor with corresponding realistic modelling variations. These tensors are then rotated within the RT-plane to investigate the influence on the global clear wood behavior. The modelled global stiffness values are subsequently compared with the experimental results.

Wood, as a naturally grown material, exhibits an inhomogeneous material structure as well as a quite complex material behavior. For these reasons, the mechanical modeling of fracture processes in wood is a challenging task and requires a careful selection of numerical methods. In this work, the focus is laid on a cohesive phase field method [1]. With this method, especially geometric compatibility issues that limit the use of, e.g., XFEM can be avoided, as the crack is not discretely modeled but smeared over multiple elements. This allows the formation of complex crack patterns, defined by the underlying differential equations and boundary conditions but not restricted by the mesh geometry. The present implementation [2,3] contains a stress-based split which allows proper decomposition of the strain energy density for orthotropic materials. Furthermore, the geometric influence of the wood microstructure on crack propagation is taken into account by a structural tensor scaling the length scale parameter of the phase field. The developed algorithm was validated on various problems. Crack patterns, including branching and merging, could be modeled very stable and accurately, even in the vicinity of knots where the material structure of wood is particularly complex and interface zones exist.

[1] J.-Y. Wu, „A unified phase-field theory for the mechanics of damage and quasi-brittle failure“, J. Mech. Phys. Solids, Bd. 103, S. 72–99, Juni 2017, doi: 10.1016/j.jmps.2017.03.015
[2] S. Pech, M. Lukacevic, and J. Füssl, „A hybrid multi-phase field model to describe cohesive failure in orthotropic materials, assessed by modeling failure mechanisms in wood“, Eng. Frac. Mech., Bd. 271, S. 108591, Aug. 2022, doi: 10.1016/j.engfracmech.2022.108591.
[3] S. Pech, M. Lukacevic, and J. Füssl, „Validation of a hybrid multi-phase field model for fracture of wood“, Eng. Frac. Mech., Bd. 275, S. 108819, Nov. 2022, doi: 10.1016/j.engfracmech.2022.108819

The growing demand for resource-efficient construction materials seeks additional engineered wood products (EWPs), focusing especially on higher utilization of the raw materials. One possibility is to use strand-based EWPs. For a broader use, a detailed characterization, as basis for a safe dimensioning is needed. This study, therefore, aims to characterize a novel EWP, in both single and three-layered configurations, highlighting its potential for construction applications. A range of standardized mechanical tests, including tension, compression, shear, and bending was carried out on both single and three-layer configurations, to identify the material’s strength and the most critical elastic constants, i.e. Poisson’s ratios, elastic moduli, and shear moduli. These properties were assessed across the material’s three principal directions, essential for accurate material modeling and dimensioning. These experimental efforts were paralleled by a first attempt at a 3D finite element model (FEM) in Abaqus to simulate the material behavior when it is subjected to out of plane bending load. In conclusion, this research not only represents a significant effort towards the accurate representation of material in mechanical properties but also a robust framework for computational models, offering promising avenues for material optimization and exploring complex loading scenarios in structural design. Future work will focus on expanding and improving the modeling framework and conducting plastic analysis and failure criterion, further aligning numerical simulations with real-world material performance.

Previous investigations demonstrated the potential of birch plywood as a valid substitute for steel plates in connections of timber structures. However, in those studies, the tests on birch plywood were limited to a simple stress state (e.g. uniaxial tension), instead of being in a complex stress state with combined multidirectional tension and compression forces, which is typical for a truss node. Furthermore, the size of the tested specimen is relatively small compared to the real case.

In this study, 6 full-scale trusses with a span of approximately 6.5 meters were tested. Each truss consisted of glulam elements with birch plywood plates at the nodes, connected by either MUF-based glue or smooth dowels with a diameter of 8 mm. The thickness of the birch plywood plates varied between 9, 12 and 21 mm.

As designed, failure should occur in the node of the truss, which was subjected to compression from the vertical post and tension from the two rafters. Plywood failure occurred in all the dowel trusses and the glued truss with 9 mm plywood, while glue line failure occurred for the glued trusses with 12 and 21 mm plywood. The doweled truss with 21 mm plywood plates will be tested at the end of April 2024. Certain disparity was observed when comparing predicted and actual failure modes. For example, plywood failure was predicted for the glued truss with 12 mm plywood, however, glue line failure was observed.

Furthermore, analytical estimations showed a slight overestimation of the experimental capacity. This is currently attributed to the fact that the nodes were assumed as hinges in the calculations, despite the fact that they are able to transmit bending moment. Lastly, the thickness of the plywood plates depicted an influence on the load-carrying capacity of the trusses but no significant influence on the stiffness.

Wood is an excellent building material, as evidenced by its remarkable strength-to-weight ratio. Designing for high reliability with an extremely low failure probability, requires a deep understanding of material properties characterized by probability distribution functions (PDFs). However, directly determining these PDFs through experiments or simulations for such low probabilities is challenging due to their dependency on geometric size and loading conditions.

The finite weakest-link theory [1] provides a practical solution for quasi-brittle materials like wood, employing a combined Weibull and Gaussian distribution. Initially, the Weibull distribution describes only the far-left tail for sufficiently small structures, while as the size increases, the distribution transitions from Gaussian to Weibull dominance. The scaling of the PDF is based on the size of a representative volume element (RVE) and the elastic stress state. Identifying the distribution parameters is challenging, requiring sufficiently large sample sizes of various structural sizes, including large-scale ones.

We estimated all the distribution parameters, allowing us to scale the PDF of bending strength for glued laminated timber beams. This was made possible by a comprehensive simulation campaign [2], where beam sections were simulated under constant bending moments, considering discrete cracks and plastic deformations. Additionally, we utilized experimental data and adhered to standards. With this framework, we obtained bending strength PDFs specific to the beam size and loading configuration. As a result, we were able to quantify the size effect and determine size-dependent failure probabilities.

[1] Z.P. Bažant and S.-D. Pang, “Activation energy based extreme value statistics and size effect in brittle and quasibrittle fracture”, J. Mech. Phys. Solids, 55 (2007) 91–131, doi: 10.1016/j.jmps.2006.05.007.

[2] C. Vida, M. Lukacevic, G. Hochreiner, J. Füssl, “Size effect on bending strength of glued laminated timber predicted by a numerical simulation concept including discrete cracking”, Mater. Des., 225 (2023) 111550, doi: 10.1016/j.matdes.2022.111550.

Timber is one of the materials that continues to be very important in the field of modern constructions. There are many researches carried out in the field of production and use of Glulam beams (GLT) and Glued solid timber (GST) and many of them are not based on the relevant standard of Eurocode 5.

The research aims to present the experimental and analytical analysis, beginning from the procedure of selection of raw material according to EN 14080, determination of moisture content of test pieces, defining dimensions of the samples (cutting slices), measuring, scaling samples and its drying. Determination of density of test pieces according to ISO 13061-4:2014 and production with all phases.

The test slice was cut of full cross section referred to standard EN 13183-1 and minimum 20mm dimension in the direction of the grain, at a point by at 300mm from either end of the test piece.

Each test slice was marked and measured in longitudinal, radial and tangential directions (8 measurements were done per each piece with Nonius Caliper) and each piece has been scaled (accurate equip. to 0,01 g) and dried in an oven dry state. Equipment for drying ensured free internal circulation of air and capable of maintaining a temperature of (103 ± 2) °C. The density of wood is determined based on the ISO 13061-4:2014 standard.

After drying the samples and determining the moisture content and determining the density of the samples experimentally, the Spruce wood (Picea Abies PCAB) was found to be of class T30 (C50) according to EN14080, material of sufficient quality for use in the production of beams. The production of Glued Solid Timber beams in this study for this experimental purpose, turned out to be a promising and well-executed for the use in constructions.