Operator splitting? What is it in the context of transport modeling?

[bumblebee77]

Hello all:

I'm a thermodynamicist wanting to get into transport modeling. In my reading, I have come across "operator splitting." I'm confused about exactly what this means in the context of transport modeling.

This is probably a silly question, but does it refer to an advective term relative to a diffusive term in a transport equation? Or does it refer to doing a thermodynamics (or equilibration/compositional) step and THEN a fluid flow step?

I have set up my own little example code to model chemical reactions during flow. I first carry out a thermodynamics step to get compositional information and then I perform a flow step. Is this operator splitting??

Thanks in advance for your patience with a rookie!

 

[anorlunda]

ping @Chestermiller . Can you help with this workshop question?

 

[Chestermiller]

Hello all:

I'm a thermodynamicist wanting to get into transport modeling. In my reading, I have come across "operator splitting." I'm confused about exactly what this means in the context of transport modeling.

This is probably a silly question, but does it refer to an advective term relative to a diffusive term in a transport equation? Or does it refer to doing a thermodynamics (or equilibration/compositional) step and THEN a fluid flow step?

I have set up my own little example code to model chemical reactions during flow. I first carry out a thermodynamics step to get compositional information and then I perform a flow step. Is this operator splitting??

Thanks in advance for your patience with a rookie!

Use of dimensionless groups is a very powerful tool for analyzing physical and chemical problems. Over the years, I have come to recognize that all physical systems are more fundamentally described in terms of their dimensionless groups than in terms of the individual parameters associated with the system. So for example, even though the total number of physical parameters involved may be 8, the number of dimensionless groups involved may only be 4 or 5. This allows the results to calculations to be presented once and for all in terms of the limited number of dimensionless groups rather than in terms of varying the individual parameters. This makes the results much easier to understand and more concise. And, if a system behavior is being examined experimentally, it reduces the number of experiments needed to quantify the range of behavior. Finally, straightforward methods have been developed for reducing the differential equations or algebraic equations for a system to dimensionless form so that the behavior can be directly determined in terms of the key dimensionless groups.