PSI - Issue 13

Uroš Tatić et al. / Procedia Structural Integrity 13 (2018) 496– 502 Author name / Structural Integrity Procedia 00 (2018) 000–000

499

4

which would directly reflect on the structural and thermal characteristics as well as the price of production and cost of finished product. A variety of the wood species are used currently on the wooden window market where Norway spruce (Picea abies), Scots pine (Pinus sylvestris), Siberian larch (Larix Sibirica), oak (Quercus spp.) and some tropical wood species are the most widely used at European market (Ugovšek A at al., 2018). Tested windows were produced of Norway spruce due to its lowest modulus of elasticity compared to other wood spices (Green, D.W. at al., 1999, SIST EN 338:2010 and Premrov M. and Dobrila P. 2008). Prefabricated wood boards were made of individual wood boards layered and bonded together. Depending of the profile it can be made of 3 to 5 layers. In the case of window simulation, wood is one of most complex and important materials to model, because it represents approximately 80 % of the window structure, has complex orthotropic behaviour, is the softest material in whole assembly (except rubber sealing) and most of the mutual contacts between elements is between wood segments. Parameters of a wood (Norway spruce) were defined as an orthotropic due to significant difference between longitudinal and transversal/radial direction. Dowels used to provide mutual contact between window elements were made of Acacia wood but were modelled as isotropic behaviour. All of the parameters used to define materials (wood species, aluminium, steel and glass) as well as their application are presented in the Table 1. 4. Experimental results Airflow, Pressure, and deformation at three locations have been monitored during the experimental setup. Experimental values for a specimen W4 and R4 (complete windows with and without reinforcement) are presented in Fig. 3. Factors of relative displacement in the middle measuring point compared to a specimen W are presented in Fig. 4. Each test was repeated in two series. Obtained results for all models have shown that the location of the highest value of deflection is at the middle of the mullion as it was expected. Deformation at the bottom and top measuring location were almost identical in the case of each loading.

Fig. 3 Experimental values for a specimen W4 (complete windows without reinforcement) and R4 (complete windows with reinforcement)

Fig. 4 Factors of relative displacement in the middle measuring point compared to a specimen W

5. Numerical simulation Development of the models was performed in couple of individual stages. Main idea behind this principle was to determine which factors (present in real experiment) have influence on the numerical simulation, and behaviour.  Model 1: Window frame model was represented as an assembly connected using tie constrain, isotropic wood behaviour was used.  Model 2: Frame elements were created as individual parts, dowels were used to restrain contacts between two main window parts, orthotropic wood behaviour was used. Applied knowledge obtained through replication of experimental testing was also used to simulate additional loading