It is probably a fact that a distributed mathematical model could represent (or model) a target system more precisely and rigorously. However, we also admit that too much detailed model could be too heavy to solve.
These days, there are many demands in CFD simulation to get a better understanding of chemical processes and the phenomenon involved. In the meantime, we also perform a traditional process simulation for design and optimization study. How come such different approaches could be integrated to make our life easy. I think we could find a solution from "hydrid modeling".
A multi-compartment modeling method (based on UNIT CELL model) could offer followings:
- Allow one to model and simulate rigorous mixing effects by complex equipment geometry
- Useful for reaction system that requiring viscous fluid dynamics simulation
- More realistic representation of industrial reactor rather than perfect mixing RBatch or RCSTR model
- Useful for reaction system that requiring viscous fluid dynamics simulation
- More realistic representation of industrial reactor rather than perfect mixing RBatch or RCSTR model
Using UNIT CELL, one could represent a geometry and also characterize "interractions" between CELLs.
This concept could be extended to much more complex modeling. And it could be coupled with CFD simulation. Well, the degree of the coupling is truely a demand-basis. From experiences, loose-coupling may b enough (i.e. few manual iteration should suffice the requirement). However, it should be possible to close the "Loop".
The above captured screenshots are from a consulting project that I provided in 2005. I still remember the customer really liked it. By the way, this modeling work has been done using Aspen Custom Modeler.
