11 Integration of Modelling at Various Length and Time Scales 231 (a) (b) Fig.11.3.(a)MesoDyn simulation of Pluronic PL85 in water at 27%concentration and temperature 15C.The isodensity surfaces of the hydrophobic component are shown,clearly revealing the micellar structure.(b)Morphology after increase of temperature to 70C,with the appropriate interaction parameter.The spherical micelles coalesce into rods,in line with experimental evidence. in the simulation was found to match well with TEM results for the real materials. A clear pathway has therefore been defined and is being more and more established for coarse-graining from the atomistic to the mesoscale.The ma- jor hole in the technology remains the reverse mapping from the mesoscale to the atomistic,where no adequate method has been developed. 11.4.2 From Mesoscale to Finite Element Simulation The structures formed on the nanometer scale give rise to diverse and in- teresting material properties.As we have seen in the last section,mesoscale methods can be used with confidence to predict such structures.While some properties can be predicted directly from the mesoscale,property predic- tion given the knowledge of material structure and the property of the pure components that comprise the mixture has been developed widely in Finite Element Methods.An example of such a method,designed to deal with finely textured materials is Palmyra-GridMorph from MatSim [11.2.Using stan- dard solvers the finite element code can then predict the property for the realistic structured material. As a test case for this combination of mesoscale and finite element meth- ods we studied the oxygen diffusion through a material designed to act as a gas separation membrane.A binary blend of polystyrene and polybutadiene was simulated with MesoDyn using parameters obtained from atomistic level modeling.These polymers tend to phase segregate and large domains form11 Integration of Modelling at Various Length and Time Scales 231 (a) (b) Fig. 11.3. (a) MesoDyn simulation of Pluronic PL85 in water at 27% concentration and temperature 15◦C. The isodensity surfaces of the hydrophobic component are shown, clearly revealing the micellar structure. (b) Morphology after increase of temperature to 70◦C, with the appropriate interaction parameter. The spherical micelles coalesce into rods, in line with experimental evidence. in the simulation was found to match well with TEM results for the real materials. A clear pathway has therefore been defined and is being more and more established for coarse-graining from the atomistic to the mesoscale. The ma￾jor hole in the technology remains the reverse mapping from the mesoscale to the atomistic, where no adequate method has been developed. 11.4.2 From Mesoscale to Finite Element Simulation The structures formed on the nanometer scale give rise to diverse and in￾teresting material properties. As we have seen in the last section, mesoscale methods can be used with confidence to predict such structures. While some properties can be predicted directly from the mesoscale, property predic￾tion given the knowledge of material structure and the property of the pure components that comprise the mixture has been developed widely in Finite Element Methods. An example of such a method, designed to deal with finely textured materials is Palmyra-GridMorph from MatSim [11.2]. Using stan￾dard solvers the finite element code can then predict the property for the realistic structured material. As a test case for this combination of mesoscale and finite element meth￾ods we studied the oxygen diffusion through a material designed to act as a gas separation membrane. A binary blend of polystyrene and polybutadiene was simulated with MesoDyn using parameters obtained from atomistic level modeling. These polymers tend to phase segregate and large domains form