The effects of moisture content and load duration on the behaviour of glued laminated timber three-pinned portal frames with tapered members

Abstract

The early 1900s were the beginning of the lengthy history of glued laminated timber's patent of the construction method was taken in Switzerland. Since then, the technics of lamination, adhesive properties and manufacturing facilities have developed, and this helped glued laminated timber to spread around the World. Glued laminated timber, which is usually treated as an orthotropic material in engineering (Mascia & Lahr, 2006), is made of two or more layers of processed lumbers parallel to each other with structural adhesives. The material's strength and stiffness differ based on the direction of the material and the direction of the loading therewithal. There are many motivations to increase the usage of timber, which are included but not limited to: being a light structural material that provides easier constructability, less seismic load due to light mass, reducing carbon emission by acting as a carbon store, being sustainable material, performing slow burn and being predictable during fire and more. The three-pinned portal frame, one of the most common glued laminated timber structures, is the subject of this thesis. The thesis aims to understand the effect of the moisture, load duration and the service classes of the glued laminated timber on three pinned portal frame members along with the glued laminated timber material's structural behaviour. The span and spacing of the frames have been chosen with consideration of commonly used construction products. Therefore, the strength classes considered are; GL24, GL28, GL30 and GL32. All service classes and the applicable load durations have been considered for the analysis. The symmetrical tapered portal frame was designed with 24 m of spanning with an assumed 5 m of frame spacing. To analyse the portal frame's maximum of 1 capacity utilisation capacity, the design of the portal frame has been determined based on the most critical material, loading term and the service class, which are GL24 class timber, medium-term loading and service class 3. There are three service classes, all of which have been included in this study. However, only two of the five load durations have been included in this study. The service classes differ from 1 to 3 depending on the material's moisture content and the surrounding environment's humidity. The class is higher where the moisture content is higher (British Standards Institution, 2004). On the other hand, the load durations are specified in EC5 according to the load type and the accumulated load duration range. For the study case, the only load that can have different load durations is snow, which the considered load durations are medium-term and short-term load durations. The analysis and design of the glued laminated timber members have been done primarily based on Eurocode standards, and where geographical information is required but not available, the Turkish standards have also been considered. So, the characteristic snow load has been taken from TSE 498 (Turkish Standards Institution, 1997) due to the lack of Turkey details in EC1-Part 1.3. The seismic load has been carried out per EC8, spectral type 1(higher seismic), and the ground and the seismic zone parameters have been chosen according to the structure's location, which is in Turkey, Marmara region. The reference peak ground acceleration is taken from the Turkey 475 return period seismic map (TMMOB Chamber of Geology Engineers, 2022). The fundamental period of vibration (T1) was first calculated with the Rayleigh method with the required fictitious of 1 kN horizontal load to be applied at the top of the columns in the Sap2000 model to determine the deflection. Additionally, the T1 period has been checked with modal response spectrum analysis to compare the results. It has been seen that both results are only slightly different, and they are both between TB and TC. So that the base shear force is found per the first period and the ordinate of the design spectrum (Sd(T)). This thesis study worked with the programs: 2-D SAP2000 analysis, AutoCAD, TEDDS for Word and MS Office. The frame members, the joint releases, and material properties have been assigned along with the specified loading and the combinations per EC0 and EC5 in the SAP2000. This software provided the moment bending, shear and axial force body diagrams in addition to the initial joint deflection. According to the first design, the tapered beams and columns are designed with 1.00 capacity utilisation capacity: the beams' thickness is from apex to the end, 0.4 m to 1.225 m, and the columns' are from pinned to the other end, 0.5 m to 1.3 m. The width of the beams and columns is 0.25 m. The details have been provided in section 4 with drawings. To run the analysis of beam and column design multiple times under different loadings and different materials strength, the calculation sheet has been prepared with the program TEDDS for Word. In these formula sheets, the most critical governing strength and deformation combinations are considered. The results showed that the ULS govern case is 1.35G + 1.5S for both columns and beams, whereas the SLS govern case is G + S for beams, and the governing case of the column is G + W. The capacity utilisation results have been gathered and plotted with MS Excel charts initially based on the material strength class, limit state, and whether it is a beam or a column. The interpretations of these graphs are provided in the conclusion, and additional graphs are included. The interpretation of results has been undertaken for ULS and SLS separately for all considered materials. Additionally, a combined ULS and SLS chart has been presented for the GL24 class material. The combined ULS and SLS graphs of GL24 material for both beam and column showed that the governing limit state is the ultimate limit state compared to the serviceability limit state with a capacity utilisation difference of more than %40. The overall governing case is the strength of the beam compared to the column with the capacity utilisation of 1.00 where the considerations are; service class 3 and load-duration class of snow is medium term. Also, one of the main interpretations is that the columns are governed by the stability capacity, whereas the beams are governed by stress capacity. It can also be seen that the ULS utilisation of both services, class 1 and 2 are equal due to the same kmod coefficient. However, this differs for service class 3 due to different load-duration classes of snow. Additionally, it is noted that the strength class of beams and columns can be selected differently for economical solutions because the capacity utilisation of columns is always less than the beams' capacity utilisation. The serviceability limit state has been calculated for portal frames' vertical and horizontal displacement limitations. During the limitation consideration, it is noted that the portal frame does not have any crane and the roof to support non-flexible or brittle parts (Steel Construction Institute, 2006). One of the findings is that the capacity usage difference between the vertical and horizontal deflections is significant in the serviceability limit state capacity utilisation. It demonstrates that vertical displacement is more critical than horizontal deflection. Vertical displacement has the highest serviceability capacity utilisation of 0.524, while horizontal displacement has 0.19. Additionally, the portal frame's vertical deflection is governed by the initial deflection, whereas the horizontal deflection limitation is governed by the final deflection. Thus, regardless of the service class difference, the SLS capacity utilisation is the same for vertical deflection. Therefore, no service class and moisture content impact can be observed on the portal frame's vertical deflection. These results can be used in future designs, making the design more efficient.