J. Cent. South Univ. Technol. (2011) 18: 2080−2084 DOI: 10.1007/s11771−011−0946−5 Computational simulation of convective flow in the Earth crust under consideration of dynamic crust−mantle interactions ZHAO Chong-bin(赵崇斌) 1 , PENG Sheng-lin(彭省临) 1 , LIU Liang-ming(刘亮明) 1 , B. E. HOBBS 2 , A. ORD 2 1. Computational Geosciences Research Centre, Central South University, Changsha 410083, China; 2. School of Earth and Environment, The University of Western Australia, Crawley, WA 6009, Australia © Central South University Press and Springer-Verlag Berlin Heidelberg 2011 Abstract: The finite element method was used to solve fluid dynamic interaction problems between the crust and mantle of the Earth. To consider different mechanical behaviours, the lithosphere consisting of the crust and upper mantle was simulated as fluid-saturated porous rocks, while the upper aesthenospheric part of the mantle was simulated as viscous fluids. Since the whole lithosphere was computationally simulated, the dynamic interaction between the crust and the upper mantle was appropriately considered. In particular, the mixing of mantle fluids and crustal fluids was simulated in the corresponding computational model. The related computational simulation results from an example problem demonstrate that the mantle fluids can flow into the crust and mix with the crustal fluids due to the resulting convective flows in the crust−mantle system. Likewise, the crustal fluids can also flow into the upper mantle and mix with the mantle fluids. This kind of fluids mixing and exchange is very important to the better understanding of the governing processes that control the ore body formation and mineralization in the upper crust of the Earth. Key words: computational simulation; crustal fluids; mantle fluids; fluid dynamic interaction 1 Introduction With the birth and development of the emerging computational geoscience discipline, computational simulation methods have been widely used to solve a broad range of geoscience problems [1]. For example, advanced numerical methods have been employed to simulate the following problems: 1) convective magma flow in the mantle [2−6], 2) convective pore-fluid flow in hydrothermal systems within the upper crust of the Earth [7−9], 3) precipitation and dissolution of minerals in pore-fluid saturated porous rock masses [10−13], and 4) fully coupled problems involving material deformation, pore-fluid flow, heat transfer, mass transport and chemical reactions within the crust of the Earth [14]. Although extensive research has been conducted to investigate the effect of convective pore-fluid flow on ore body formation and mineralization within the upper crust of the Earth, little work has been done to understand the effect of dynamic crust−mantle interaction on the convective pore-fluid flow within the crust of the Earth. It has been widely recognized that mantle convection is the primary mechanism for the transport of heat and mass from the Earth’s deep interior to its surface [6, 15], so that it influences the Earth’s topography, gravitational field, climate system, cycles of glaciation, biological evolution, and formation of mineral and hydrocarbon resources. As a result, mantle convection becomes the fundamental driving force of plate tectonics, formation and drift of continents, volcanism, earthquakes, and mountain building. Since the thickness of the mantle (i.e. a few thousands of kilometers) is much greater than that of the crust (i.e. only a few tens of kilometers), the crust acts as the skin of a huge body in the whole crust−mantle system. From the computational simulation point of view, this large difference in thickness between the mantle and the crust creates a severe difficulty, when the whole crust−mantle system is modeled simultaneously in a computational simulation. For instance, if the finite element mesh is designed to produce a useful solution for the mantle, then it cannot give any meaningful solution to the crust, because the mesh scale is too large to model the details of the crust. On the other hand, if the finite element mesh is designed to simulate the detailed phenomena within the crust, then it may become computationally impractical because a huge number of degrees-of- freedom are created to model the mantle. To overcome the aforementioned difficulty, current numerical practice is to simulate the crust and mantle Foundation item: Project(10872219) supported by the National Natural Science Foundation of China Received date: 2010−10−09; Accepted date: 2011−01−21 Corresponding author: ZHAO Chong-bin, Professor, PhD; Tel: +86−731−88830039; E-mail: chongbin.zhao@iinet.net.au