mesnardo
commented
8 years ago - Modify title of Story 2 to remove the word "trap". See change history.
- Polish language in intro about IBAMR. See change history.
- Fix imprecisions in description of IBAMR. See change history.
- Move up the cite to Bhalla et al. 2013. See change history.
- Polish description of outlet BC with IBAMR. See change history.
- Polish language about interior body constraint. See change history.
- Explain sub-figure with stabilization. See change history.
- Add mention of improved result with lower CFL. See change history.
- Refine language in story-2 postmortem removing negative connotations. See change history.
- Modify mention of negative results. See change history.
Some of the language in the section reporting our experience with IBAMR can be read as direct criticism towards the team that developed this library. We've heard from the IBAMR team that they object to our description.
In particular, the sentence "The published record is incomplete in this regard: we could find no explanation for it [the use of a solid body discretization instead of a boundary discretization] in any paper using IBAMR" was interpreted as an accusation that the papers deliberately leave out important details. This was not our intention in writing the text. We simply want to report our experience trying to learn how to use this code, and our mistakes in the process.
We also need to correct the description of the outflow boundary conditions, which are not zero-gradient conditions, but traction-free conditions. This was our mistake in the description. And as the discussion following the post in our group web site shows, the IBAMR team is worried that readers could misinterpret Figure 5 as "a fundamental problem with the library." We need to edit the figure caption to make clear that it was our error setting the boundary condition, at first, and that we corrected it later. The section does say we successfully replicated previous results with IBAMR, but it can be said also in the caption.