It's interesting that you are looking into this and we welcome any feedback. Feel free to post the results of your analysis on the blog or wiki or just the links.
The results should probably be the same as any standard FV or FD treatment of the same problem. I have no idea what could cause the discrepancy. Good luck with it though. On Fri, Mar 23, 2012 at 11:59 AM, Fausto Arinos de A. Barbuto < [email protected]> wrote: > > Hello, > > I observed that FiPy's outputs for Example 5.1 (mentioned in an earlier > note I sent > to the list) are more accurate than the results printed in the book, > results which I > was able to accurately obtain with an Octave script I wrote in the past. > In fact, FiPy's > outputs are way closer to the analytical solution than the results > presented by > Versteeg & Malalasekera, regardless of the method I selected to handle > the convection > term (I've tried both PowerLawConvectionTerm and ExponentialConvectionTerm > with > quite similar results). > > Where's the black magic? ;-) How can FVM be more accurate than FVM itself? > > Fausto > > > > ------------------------------ > *From:* Fausto Arinos de A. Barbuto <[email protected]> > *To:* FiPy <[email protected]> > *Sent:* Friday, March 23, 2012 11:50:32 AM > *Subject:* Matrix display > > > Hello, > > To solve a steady-state diffusion-convection problem such as the one > presented in > pg. 83 of FiPy Manual Release 2.1.3 > (examples.convection.exponential1D.mesh1D) > a tridiagonal and a column matrix must be put together. I was wondering > if there > isn't a way to display these matrices using some FiPy built-in function. > Is that > possible? I did a quick (and unsuccessful) search in the manual but > didn't find > anything specifically related to that. > > I ask that because I've recently started to learn FiPy (and I'm very well > impressed > so far) and decided to solve the worked-out examples in Versteeg & > Malalasekera's > great book, "An Introduction to Computational Fluid Dynamics". I know, > the > examples are very simple but, as I said, I've just started with FiPy and > the idea is > to move on to more complex situations later. I successfully solved the first > two pure > diffusion examples (4.1 and 4.2) and decided to tackle a diffusion-convection > one, > namely Example 5.1 in page 137. The results I found with FiPy are very close > to the > figures presented in the book, but aren't quite the same. That is the > reason why I > want to inspect the matrices -- I want to compare them with the ones > presented in > the book. I'm sure this is not a FiPy problem, it's my code. Hopefully > the matrices > will tell me where I'm goofing. > > Thanks, > > Fausto > > _______________________________________________ > fipy mailing list > [email protected] > http://www.ctcms.nist.gov/fipy > [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ] > > > > _______________________________________________ > fipy mailing list > [email protected] > http://www.ctcms.nist.gov/fipy > [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ] > > -- Daniel Wheeler
_______________________________________________ fipy mailing list [email protected] http://www.ctcms.nist.gov/fipy [ NIST internal ONLY: https://email.nist.gov/mailman/listinfo/fipy ]
