Morio TOSHIMA Reverse analysis of soil properties based on In-situ data obtained by articureted shield II Supervisor Mitsutaka SUGIMOTO Today, the control and operation of shield is carried out by using an automatic shield control system. But ground properties, acting force on shield, and shield behavior during excavation are not clear. Therefore, the shield kinematic model had been developed. The validity of this model was verified by using site data with various conditions. But the ground properties, which are determined by site investigation reports and empirical equations, has an wide range sometimes. To obtain the ground properties relating with ground reaction force precisely, in this study, the reverse analysis method was developed to obtain the ground properties from jack force, shield position, and shield postulate, and its performance was validated by using in-situ data. The tunnel at the test site was excavated by an articulated shield in Alluvial very soft clay with SPT-N value 0. The "articulated test" was carried out to measure the shield behavior and the required crease jack force when the shield stopped at an straight alignment and the articulated angle increased step by step. As a result, the followings were observed: 1) the shield rear body rotates to the opposite direction of the shield front body rotation direction; and 2) the crease jack force increased linearly as the articulated angle increased. In the analysis, at first, the properties concerned with the shield operation were identified by the simulation on shield behavior using the in-situ data just before the test section. Based on the above obtained properties, the reverse analysis on the ground properties, such as coefficient of earth pressure at rest KH0, coefficient of active earth pressure KHa, KVa, coefficient of passive earth pressure KHa, KVa, and coefficient of subgrade reaction kH, kV, was carried out applying the direct search method to the equilibrium conditions of horizontal moment, which is considered to be most influenced by horizontal crease, at front body where the influence of tail force is negligible. As a a result, the followings were made clear: 1) the obtained value of KH0, KHa, KVa, KHa, and KVa is almost same as that of the existing empirical equations; and 2) the obtained value of coefficient of subgrade reaction, kH, kV is 3 or 4 times, compared with that of the existing empirical equations. This is considered as follows: 1) since the existing empirical equations on coefficient of subgrade reaction were obtained from the measured data including the influence of ground loosening, these equations will underestimate the coefficient of subgrade reaction of intact ground; 2) the borehole loading test will underestimate the coefficient of subgrade reaction due to the scale effect.