In the free flow, it is reasonable to assume the particle has a very small delta U with the water. In the boundary layer where the water is changing directions, the relative horizontal velocity can change dramatically...
I am not confident this is implemented in a physically consistent way, so we could use some discussion about the simplifying assumptions of the Lamb model and whether we've broken them by adding physics. I don't recommend merging yet.
Also, fixed an underflow bug that ocurred when wrel was very small and then that very small number was squared. A likely source of our nondeterministic behavior and a definite cause of the "Nan in pi_x" bug that I had trapped but had not yet located. I will separately patch this fix to main.
In the free flow, it is reasonable to assume the particle has a very small delta U with the water. In the boundary layer where the water is changing directions, the relative horizontal velocity can change dramatically...
I am not confident this is implemented in a physically consistent way, so we could use some discussion about the simplifying assumptions of the Lamb model and whether we've broken them by adding physics. I don't recommend merging yet.
Also, fixed an underflow bug that ocurred when wrel was very small and then that very small number was squared. A likely source of our nondeterministic behavior and a definite cause of the "Nan in pi_x" bug that I had trapped but had not yet located. I will separately patch this fix to main.