Messages
12.00 Sujay Deshpande
12.30 Marius Holm
13.00 Iver Frimannslund
13.30 Kai
14.00 Mattias Rauter
14.30 Yiyi Witchelo
15.00 Lars Molstad
The exam will be held after lunch on Thursday. A more detailed schedule will be posted on Wednesday.
For the presentation, you have 15 minutes for the final project followed by 15 minutes of questions from the syllabus. Power-point, Keynote, PDF or anything else that works is ok. I will have a Mac hooked up and you can either send me the presentation, bring a usb or your own laptop.
The presentation should start with a project description. It is difficult to give general advice, but briefly it is your responsibility to make the presentation as clear and concise for the external sensor as possible. Emphasize what you have done, both why and how you have done it the way that you have. Explain the results. Model sensitivity, mesh sensitivity and sensitivity to numerical methods are relevant topics for most projects.
and will not return until Monday 3/6. So if you have any questions about the project or final exam please come tomorrow at regular hours. Note that I'll still be available on email all the way up to the exam.
Final oral exam will be in the first week of June. Please let me know asap if there are some dates that are not possible due to other exams. Will fix final date tomorrow.
Jeg m? p? kurs b?de onsdag og torsdag. Jobb med prosjektet. Dersom du ikke allerede har f?tt prosjektbeskrivelsen godkjent ta kontakt.
This week you should present to me a project description like we talked about 10/4. Maximum two pages. The project description must be approved no later than next week.
I'm not at UiO today, but send me an email if you have questions about the project.
You should now start to think about the project. I will be available both Wednesday and Thursday for discussing ideas. Next week a final project description must be submitted.
Implement a backwards facing step case in OpenFOAM. Use the icoFoam solver, an expansionfactor of 2 and Re=200 (laminar) and Re=1000 (unstable, turbulent?). See, e.g., this link
For Re=1000 the default numerical scheme in OpenFoam should result in an unstable solution and wiggles. Find an appropriate scheme that removes the wiggles (upwind/TVD).
Consider Example 5.1 in Versteeg et al. Implement a solver using TVD for convection and show convergence in the L2-norm. Choose any or all of the TVD schemes from the Table on p170. As a suggestion, but not a requirement, try to implement a generic TVD solver where the scheme can be chosen from the command line.
convdiff.py