Abstract
Wood has been used as a structural material since ancient times, so we have always had to face the challenges of time and climate on wood and wood-based materials. In those situations, it expresses the creep behavior, a mechanical phenomenon in which the deformation increases with time at a constant level of load.This study started from the overview of wood material and the construction of a database collecting wood creep measurements, representing the diversity of creep test methods and results from the literature. In the present study, experiments were focused on four species representative of the French resource : Douglas fir, Poplar, European Beech and European Oak. We chose a dedicated 4-point bending test for creep measurements associated with vibration procedure for pre-characterization of mechanical properties, as well as the physical measurements of density and dimensions. The clear wood samples were small (150 mm in the direction of the fibers) to avoid wood defects.The vibration testing device has been designed at LMGC based on a system developed in Japanese wood laboratories. It allows us to measure the specific modulus, which is strongly correlated to the microfibril angle and the damping coefficient, in order to have a preliminary understanding of the mechanical characteristics of the samples. The density, specific modulus, damping coefficient, volume swelling coefficient, and grain angle of the specimens were measured and considered into an optimized sampling. The sampling method was developed to choose a specific property as a variable and control the variance of the other properties, so that the test results can focus on the effect of a given property. Thus, the influence of each property is tested separately from the others to determine the predictive indicators of creep.Several series of creep tests were conducted to study the effect of each predictive indicators. All the experiments were conducted in controlled environment at temperature 20 °C and relative humidity 85% after a stabilization phase of the specimens, then the 4-point bending creep tests were conducted. A load of 650 g (10.52 MPa) was set on the specimens for a nominal duration of 10 days. There were 108 samples tested and considered in the modeling.Creep behavior model can be expressed by an exponential function of time. Delayed compliance can be represented by a series of Kelvin elements, and the parameters of the elements (springs and dashpots) are calculated numerically by fitting with the experimental data. Different from the traditional model, this study focuses on the relationship between the fitting parameters and the specimens characteristics. Three models are presented to estimate the fitting parameters of the creep function. The modeling suggests that the creep behavior of wood occurs in the cell wall rather than by sliding at the inter-cell zone, and the mechanism changes with time are shown in the modeling result. The effect of density decreases with the loading time, and the effect of specific modulus increases. It means that the deformation is transferred from the cell wall to other structure of wood, and the influence of microfibril angle increases during the loading process.This study highlights the importance of microstructure on rheological behavior. Wood microstructure is influenced by age, growing environment and conditions, as well as wood extractive. The results of this study provide material properties to be considered in the mechanical grading assessment of construction materials.