Stabilization of subgrade is an internationally common technique to improve the characteristics of a soft soil. Improving the stiffness and bearing capacity of the subgrade enables the pavement to be designed with reduced depths of subbase and basecourse aggregates. This makes the stabilization technique both economically and environmentally interesting. Both natural and stabilized soils have high inherent property variability that should be considered when the characteristic value is derived. By analyzing the spatial variability in stabilized and natural subgrades it is possible get the parameters needed for reliability–based modeling of soil properties. I order to do so, three statistical parameters need to be evaluated, i.e. the mean, the variance and the scale of fluctuation, which is the distance within which soil properties show relatively strong correlation. The objective of this thesis is to study different devises for determining the elastic modulus of subgrade and to study the spatial variability of the elastic modulus. With knowledge of this it is possible to optimize the testing. The different devises used in the field investigations was the light weight deflectometer (LWD), surface seismic and static plate loading. The different devices were used because they take measurements at different load- and strain levels. Another aim was to investigate correlations between moduli determined by the different devices. This thesis is a part of a research project at Skanska about subgrade stabilization. The field studies were performed in Tygelsjö, south of Malmö, and the soil type was claytill. Two test areas, each 7x15 m, were stabilized with 6 % respective 4 % hydrated lime and two test areas, each 4x15 m, were left unstabilized. Totally 133 test points were investigated with the LWD and surface seismic and 11 static plate load tests were performed. With geostatistical analyses of the experimental results the scale of fluctuation with respect to the elastic modulus has been evaluated to approximately 1,3 m. That means only points within 1,3 m reveal relative strong correlation. In practice this means that subsidence due to less stiffness in the stabilized subgrade will not be extensive in the surface. Therefore it’s not economically rational to design the pavement with respect to small areas with low stiffness in the subgrade. Comparison between the LWD and the static plate load test showed that ELWD is on average 2 times greater than EV2. No correlations with the surface seismic were found. One influencing factor on the LWD measurements was the drop height. ELWD decreased with increasing drop height but only on the unstabilized subgrade. This thesis also shows how the characteristic value of the elastic modulus could be evaluated with statistical methods where the evaluation will depend on the number of measurements, the variability of the measurements and the variance reduction factor which in turn depends on the scale of fluctuation and the size of the mechanical system. This type of evaluation could be used as basis for reliability based design and design according to the observational method according to Eurocode 7. / <p>QC 20100706</p>

<p>Stabilization of subgrade is an internationally common technique to improve the characteristics of a soft soil. Improving the stiffness and bearing capacity of the subgrade enables the pavement to be designed with reduced depths of subbase and basecourse aggregates. This makes the stabilization technique both economically and environmentally interesting. Both natural and stabilized soils have high inherent property variability that should be considered when the characteristic value is derived. By analyzing the spatial variability in stabilized and natural subgrades it is possible get the parameters needed for reliability–based modeling of soil properties. I order to do so, three statistical parameters need to be evaluated, i.e. the mean, the variance and the scale of fluctuation, which is the distance within which soil properties show relatively strong correlation.</p><p>The objective of this thesis is to study different devises for determining the elastic modulus of subgrade and to study the spatial variability of the elastic modulus. With knowledge of this it is possible to optimize the testing. The different devises used in the field investigations was the light weight deflectometer (LWD), surface seismic and static plate loading. The different devices were used because they take measurements at different load- and strain levels.</p><p>Another aim was to investigate correlations between moduli determined by the different devices. This thesis is a part of a research project at Skanska about subgrade stabilization. The field studies were performed in Tygelsjö, south of Malmö, and the soil type was claytill. Two test areas, each 7x15 m, were stabilized with 6 % respective 4 % hydrated lime and two test areas, each 4x15 m, were left unstabilized. Totally 133 test points were investigated with the LWD and surface seismic and 11 static plate load tests were performed.</p><p>With geostatistical analyses of the experimental results the scale of fluctuation with respect to the elastic modulus has been evaluated to approximately 1,3 m. That means only points within 1,3 m reveal relative strong correlation. In practice this means that subsidence due to less stiffness in the stabilized subgrade will not be extensive in the surface. Therefore it’s not economically rational to design the pavement with respect to small areas with low stiffness in the subgrade.</p><p>Comparison between the LWD and the static plate load test showed that ____ is on average 2 times greater than ___. No correlations with the surface seismic were found. One influencing factor on the LWD measurements was the drop height. ____ decreased with increasing drop height but only on the unstabilized subgrade.</p><p>This thesis also shows how the characteristic value of the elastic modulus could be evaluated with statistical methods where the evaluation will depend on the number of measurements, the variability of the measurements and the variance reduction factor which in turn depends on the scale of fluctuation and the size of the mechanical system. This type of evaluation could be used as basis for reliability based design and design according to the observational method according to Eurocode 7.</p> / QC 20100706