||In the research project “ForBAU – The Virtual Construction Site” [Bormann et al. 2009], an integrated 3D model-based framework for simulating earthwork processes has been developed. This simulation framework consists of two major parts: an earthwork modeling and assessment tool ForBAU Integrator which integrates road, subsoil and terrain model in a holistic model and provides high-accurate quantity takeoffs of the earth movement as simulation source data [Ji et al. 2009]; The second important part is a discrete event-based simulation tool used to model and simulate earthwork processes in detail [Wimmer et al. 2010]. Both tools communicate via an XML interface. In addition of generating quantitative simulation source data, the ForBAU Integrator has been extended by the ability to model earthwork optimization problems using bipartite graphs and to solve these problems with linear programming techniques [Ji et al. 2010]. One of these problems is the Earthwork Allocation Problem (EAP). Here, the optimization objective is determining the optimal assignments of cut to fill areas such that minimal transportation costs incur. Beyond these research results, this paper presents a bidirectional coupling concept between microscopic simulation and macroscopic earthwork optimization. The aim of this approach is iteratively increasing the accuracy of the simulation results in the entire framework. On the one hand, the exactly optimized cut-to-fill assignments are subjected to the mathematical equations which describe upper and lower capacity limitation of respective cut or fill areas as well as the efficient distance between them. This optimization result implies the most effective way for the earth movement from a global point of view, regardless of any resource restrictions. On the other hand, the resource assignments and processing details are modeled on the simulation platform. The actual transportation time of the cut-to-fill earth movement (excavation, load, transportation, compaction, etc.) can be closely estimated in the simulation environment. The key issue behind this concept is to define an iterative parameter exchange between the two different subsystems. Obviously, two possible coupling-parameters can be applied in this case: the cut-to-fill assignments and the corresponding processing time. The iteration starts by initializing the cost function with an average processing time. The optimizer calculates based on this trivial cost function the optimal cut-to-fill assignments as input data for the simulator. After the simulation is finished, the simulated processing times will be re-imported for updating the cost function in the optimizer. The iteration process should end with a converging processing time of earthwork movements. This concept is already implemented and will soon be evaluated by means of a federal high-way construction project in Germany.REFERENCESA. Borrmann, Y. Ji, I-C. Wu, M. Obergrießer, E. Rank, C. Klaubert, W. Günthner: ForBAU - The Virtual Construction Site Project. In: Proc. of the 26th CIB-W78 Conference on Managing IT in Construction. Istanbul, Turkey, October 2009. Y. Ji, A. Borrmann, E. Rank, J. Wimmer, W. Günthner: An Integrated 3D Simulation Framework for Earthwork Processes. In: Proc. of the 26th CIB-W78 Conference on Managing IT in Construction. Istanbul, Turkey, October 2009. J. Wimmer, Y. Ji, T. Horenburg, A. Borrmann, W. Günthner, E. Rank: Evaluation of the 3D Model-based Earthwork Process Simulation in Practice. In: Proc. of the 14th ASIM-Conference Simulation in Production and Logistic (ASIM2010), Karlsruhe, Germany, October 2010, (to appear).Y. Ji, F. Seipp, A. Borrmann, S. Ruzika, E. Rank: Mathematical Modeling of Earthwork Optimization Problems. In: Proc. of the International Conference on Computing in Civil and Building Engineering 2010 (ICCCBE 2010), Nottingham, UK, Juli 2010.