You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Building on what @jm-c said, I think we want to design an interface in which an equation is expressed as something like
φ_t = Lφ + Ν(φ)
where φ is the solution (an array of fields in this case, probably), L is a linear operator that can be treated implicitly (depending on the time-stepping scheme), and N(φ) is a nonlinear operator.
The challenge is including the pressure correction in this paradigm. I think it should be possible --- we may just have to consider only time-steppers that invoke the special pressure correction substep.
I think the design should be about more than just 'marking' the terms. We can tackle the problem of designing a time-stepper to step forward an equation of a certain form completely independent from the development of an algorithm to calculate nonlinear terms, pressure corrections, and linear terms for a specific kind of equation. Then we put the pieces together to solve the problem.
The text was updated successfully, but these errors were encountered: