GeoCorner

Author: Minna Laitinen
Supervisor: Professor Leena Korkiala-Tanttu
Advisors: M.Sc. (Tech) Jani Lepistö, M.Sc. (Tech) Henry Gustavsson
Funding: Insinööritoimisto Lepistö Oy

Abstract:

In this thesis a variable head field test was studied in determining hydraulic conductivity, i.e., k-value, of unsaturated soil. Of interest was, how the test results relate to other methods of defining k-value. The primary area of interest was how useful the field test is in supporting stormwater infiltration design compared to other methods. The methods used in comparison with the field test included computational methods, laboratory tests, and literature values. Some other methods of determining the k-value are mentioned as well, although they are not used in this study. In addition, a brief survey based on expert interviews was made of the current practices of determining the k-value in stormwater design. 

Computational methods are based on the grain size of the soil. They also variably take other factors influencing k-value into account. As a result of this study, it turned out that for unsaturated soil, the formulas generally overestimate hydraulic conductivity, or are otherwise unreliable in estimating it.

Laboratory tests were performed as variable and constant head tests. In this study, the natural density of the soil was not determined, so the samples were compacted using a certain number of blows of the compactor. This probably led to samples differing in density from the natural state. Without in-situ density determination, these tests are not a reliable way to obtain natural k-values.

The results of the field tests were calculated using two formulas, one of which is simpler and developed for saturated soil, and the other for evaluating the k-value of unsaturated soil. It was concluded that the results obtained with the last-mentioned formula are probably the closest to the real k-value. The method seems to be suitable at least for sandy and gravelly soils.

Keywords:  hydraulic conductivity, variable head test, k-value, unsaturated
soil, field test

Author: Iiro Lindgren
Supervisor: Professor Wojciech Solowski
Advisors: M.Sc.(Tech) Antti Perälä (SSAB Oy), M.Sc. Petri Tyynelä (Ramboll Finland Oy)
Funding:  SSAB oy, Scandia Steel

Abstract:

All types of piles have some form of initial curvature, and it is impossible to install them completely straight. This curvature can lower the pile's structural resistance, and thus, designers need to take the curvature into account in the design. This is done using table values in standards that predict the pile's initial deflection from a straight center line.
This Master's thesis compares table values for pile curvature from Finland's RIL 254-2016 and upcoming Eurocode revision to actual pile measurements. Actual steel pile curvature measurements were collected from Finland and Sweden, and datasets were compared with the table values to determine how accurately they predict the curvature of piles. In addition to these comparisons, based on the datasets aim was to find the impactful variables affecting pile curvature. For this, two hypotheses were presented: "Jacked piles are prone to be most curved" and "Larger diameter piles are generally straighter." 
The results show that overall, the table values from Finland's RIL 254-2016 and Eurocode predict the pile curvature relatively well. Data used in this thesis agree with the hypothesis that jacked piles are prone to be more curved on average than other installation methods. However, the hypothesis that small-diameter piles are, on average more curved could not be proven with the data available in the thesis. Some parts of the data agreed with the hypothesis, and some contradicted it. Due to the problematic nature of the data, the results are inconclusive, and more detailed measurements with all background variables would be needed to answer the problems presented in this thesis more thoroughly.

Keywords:  Steel pile, curvature, initial deflection, data analysis, standards