Felix Hegemann

Hybrid Ground Data Model for Interacting Simulations in Mechanized Tunneling

Because of the ongoing expansion of urban areas, transportation systems are being increasingly moved beneath the ground into tunnels. However, the construction of tunnels is very challenging regarding unexpected geological conditions, the protection of existing buildings from damage, and the control of financial and safety risks. During the tunnel construction process, different simulations are performed to predict, for example, settlements, soil excavation, or grouting behavior. Although these simulations should strongly interact, they often use independent heterogeneous geological data models and resources, thus resulting in unreliable predictions. This paper presents a hybrid ground data model along with necessary access concepts to support the automated management and exchange of continuously updated geological information among interacting simulations. The ground data model uses a boundary and a decomposition representation of the ground geometry with material information attached to each representation. These representations are converted to one another on demand to make up-to-date ground information available for participating tunneling simulations. To validate the proposed model and concepts, a case study is presented in which two typical simulations are coupled using the prototypically implemented hybrid ground data model.

A Hybrid Ground Data Model to Support Interaction in Mechanized Tunneling

Tunneling processes are governed by diverse types of activities such as tunnel boring machine simulation, advance reconnaissance, or settlement detection. All these processes require detailed ground information on different levels of spatio-temporal scales. Therefore, a consistent ground data model is necessitated to store ground information of different categories such that diverse analyses and simulations can be provided. Due to the requirements posed by these analyses and simulations, a ground data model must be developed as a hybrid model to enable processing data in terms of boundary representation models as well as voxel-based octree models. To this end, the present paper gives an overview on how to create a ground data model tailored to the needs of mechanized tunneling. Also described is the process of reading data from an unmodified ground data model to handle tunneling specific queries. Finally, an outlook is given on future improvements.