Timber Pile-Supported Embankments : Arching and Reinforcement

Sammanfattning: Reduced climate impact is a worldwide strive today. The foundation engineering industry is continuously searching for more sustainable solutions to reduce resource usage and pollutions directly or indirectly. One such solution is timber piling, as an alternative to the commonly used concrete and steel piles. Geosynthetic-reinforced pile-supported embankment (GRPSE) is a common foundation method for settlement reduction of both roads and railways on soft subsoil. Pile-supported embankments rely on arch formation within the embankment material, which transfers the traffic and embankment load onto the piles. Reinforcing of the embankment with geosynthetics further increases this load transfer, whilst also stabilising the formed arches. Substituting concrete and steel piles with timber piles allows for a GRPSE solution with lower carbon footprint, especially if the timber piles are untreated and the (concrete) pile caps are excluded. The lower strength of timber piles and exclusion of pile caps require narrower pile spacing and/or more extensive geosynthetic reinforcement to maintain a stable arch formation in the embankment. Unstable arches can cause unwanted differential settlements in the upper structure of the embankment. Although timber piling is being practiced in countries like United States, Canada, Australia, and the Netherlands, only Sweden has a dedicated code for GRPSE using untreated timber piles. However, the Swedish code is deliberately conservative with narrow required pile spacing and two layers of GR. The aim of the thesis is to improve the resource efficiency and sustainability of geosynthetic-reinforced timber pile-supported embankments, by optimizing the required number of piles and amount of geosynthetic reinforcement (GR) based on the Swedish code.First, a numerical study was performed to evaluate the Swedish code with a focus on the pile group. The code states that the timber piles should be installed with a centre-to-centre distance of 0.8–1.2 m in a triangular arrangement instead of the more common square arrangement. Finite element (FE) modelling setups—with square and triangular pile arrangements with varying centre-to-centre distance—were used based on a geosynthetic-reinforced timber pile-supported road embankment to evaluate the design criteria. As part of the evaluation, a state-of-the-art study was done on international design guidelines and analytical models. From the FE simulations, no evident difference of mechanical behaviour is found between the triangular and the square piling pattern. The maximum allowed centre-to-centre distance between timber piles can be increased from 1.2 to 1.4 m, decreasing the number of timber piles by as much as one-third.Second, a field study was carried out. A reconstructed road embankment with geosynthetic reinforcement and timber pile support was instrumented in 2020, and the first two years of post-construction data was analysed. Monitoring of the embankment included settlements, pile deflection, pore water pressure, load on piles and subsoil, and strains in the GR. Only small strains were observed in the GR because of minor GR deflections. Partial arch formation was found from the measured load distribution, as less than half of the total load of embankment and traffic was carried by the piles. The pile loads increased in winter as the frost front penetrated the embankment and stiffened the embankment fill. The field study data provides a detailed reference for this thesis and further research on timber pile-supported embankments.Third, a numerical study was performed on the effect of geosynthetic reinforcement in timber-piled embankments. Three different cases were studied: two layers of GR (“beam” theory load transfer), one layer of GR (“catenary” load transfer) and unreinforced. The hypothesis is that there exists a range of embankment heights and centre-to-centre pile spacings for which one of the cases is preferrable to the other two in terms of resource efficiency. FE simulations, calibrated using the data from the field, were performed of the three cases with varying pile centre-to-centre spacing, embankment height and subsoil material. Based on the results of the FE simulations and literature on arch formation, limits for the three cases are suggested as a guideline for stable arch formation in timber-piled embankments. The limits allow for a more object-specific design than the Swedish code, improving the resource efficiency both in terms of required number of piles and amount of GR.Based on the findings from the three sections presented in this thesis, recommendations are given on timber pile-supported embankments. The recommendations are based on theoretical calculations and field test data. As an outlook, the thesis outlines the design of a physical laboratory test setup to verify the theoretical results and establish implementable recommendations. Though the primary application of timber pile-supported embankments, the physical test results can be extrapolated to pile-supported designs in general.