COMPUTER MODELLING OF CHEMICAL AND THERMAL DIFFUSION IN MAGMATIC SYSTEM

Abstract

Previous laboratory studies have been used to measure multicomponent chemical diffusion reactions along with diffusivities in melts of contrasting composition and to the water-saturated eutectic granite composition in an n-component system or in this case an 5-component system (K20-Na2O-A1203-SiO2-H20) at 1.0 GPa and temperatures ranging from 1300°C to 1600°C. The measured diffusion profiles are results of difference in concentrations of various chemical components, mainly oxides. However diffusive behaviour of a particular chemical component is not similar to its characteristic behaviour if there is presence of other components. The dependence of diffusion on diffusivities of other components lies in the Eigen vector, the direction dependent component of diffusivity matrix. This can be showed with a computerized model with efficient use of dynamics and mathematics behind this entire process. Diffusion always leads to small to large scale transfer of mass creating gradual concentration gradients, which vary with time. Such partitioning in natural systems would continue until concentration gradients are nullified by continued, much slower inter-diffusion. Additionally when, temperature hiatus is present, the effect of much faster but effective thermal diffusion on the above process can certainly not be ignored. This can be studied with a numerical difference approach.