(no subject)

[brooksmoses]

So, now that I'm finished with the conference, I've been working on planning out all of the remainder of the programming and computations that I need to do in order to finish my dissertation. Today, I went through and made a detailed list of the steps, so as to have an idea of how much effort this was actually going to take.

The list: (Yeah, I know it's a bit cryptic.)

2-D cylinder calculations in OpenFOAM. (Purpose: confirm that OpenFOAM is accurate; input for next step.)

• create mesh with solid-wall, symmetry, and in/out boundary conditions.
• calculation and output of secondary field variables: convection, etc.
• run calculations.
• check calculations for reasonableness.

Loading results from 2-D cylinder calculations. (Purpose: input for next step.)

• reading grid positions from grid file.
• reading field variables from field files.
• converting to appropriate representation for filtering.
• calculating surface force distribution for convergence tests.
• converting and writing out appropriate representation for plotting.

Filtering results from 2-D cylinder calculations. (Purpose: baseline for viscous-dissipation comparison; input for next step.)

• determine appropriate terms to calculate.
• calculating filtered terms.
• calculating secondary terms from filtered terms.
• calculating results for comparison: wake defect, dissipation, etc.
• output of results for comparison.
• output of fields in a format for plotting and later processing.
• reasonableness confirmation on results.

2-D Model field processor. (Purpose: input for next step.)

• input of fields from filtering processor.
• simplification of models as appropriate.
• conversion to appropriate new grid.
• output of fields in format for input into OpenFOAM.

2-D filtered calculations in OpenFOAM. (Purpose: well-resolved exact-model calculations.)

• create mesh with appropriate boundary conditions.
• check that forcing term produces reasonable results.
• calculation and output of viscous dissipation, etc.
• run calculations.

2-D postprocessing of OpenFOAM output. (Purpose: conversion to usable results; quantitative comparisons, etc.)

• reading grid positions and field variables from OpenFOAM files.
• computation of net results: wake defect, net dissipation, etc.
• output of net results in format for comparison.
• output of fields in a format for plotting.

2-D postprocessing of old output. (Purpose: comparisons of viscous dissipation for inexact models.)

• computation of viscous dissipation.
• output of net results in format for comparison.
• output of fields in a format for plotting.

3-D sphere calculations in OpenFOAM. (Purpose: exact calculations for producing models, comparison, etc.)

• create mesh with solid-wall, symmetry, and in/out boundary conditions.
• calculation and output of secondary field variables: convection, etc.
• run calculations.
• check calculations for reasonableness.

Loading results from 3-D sphere calculations. (Purpose: input for next step.)

• reading grid positions from grid file.
• reading field variables from field files.
• converting to appropriate representation for filtering.
• calculating surface force distribution for convergence tests.
• converting and writing out appropriate representation for plotting.

Filtering results from 3-D sphere calculations. (Purpose: exact results for comparison; finding model terms.)

• determine appropriate terms to calculate.
• calculating filtered terms.
• calculating secondary terms from filtered terms.
• calculating results for comparison: wake defect, dissipation, etc.
• output of results for comparison.
• output of fields in a format for plotting and later processing.
• reasonableness confirmation on results.

3-D Model field processor. (Purpose: input for next step.)

• input of axisymmetric fields from filtering processor.
• simplification of models as appropriate.
• conversion to appropriate new grid, going from axisymmetric to 3-D.
• output of 3-D fields in format for input into OpenFOAM.

3-D filtered calculations in OpenFOAM. (Purpose: calculate the actual results of the models; generate main results.)

• create mesh with appropriate boundary conditions.
• calculation and output of viscous dissipation, etc.
• run calculations.

3-D postprocessing. (Purpose: evaluation of results, conversion to usable form.)

• reading grid positions and field variables from OpenFOAM files.
• computation of net results: wake defect, net dissipation, etc.
• output of net results in format for comparison.
• slicing of appropriate field cross-sections for plotting.
• output of sliced fields in a format for plotting.

My guess is that that's probably going to take somewhere between a half-day and a day of work for each bullet point, on average. Some of them are near-duplicates between the 2-D and 3-D, which means that a lot of it can be reused from one to the other, but the reuse is still not trivial.

I'm trying to not stress about how many bullet points there are.

Comments

[larksdream]

Okay, I count more or less 60 bullet points. That means you'll be done three months from now even allowing for things to go awry! To *me* that sounds very exciting. :)

[kalmn.livejournal.com]

that's a fair bit of stuff. but you can do it.