search

yifanzhu-fluid / PCOMCOT

An efficient parallel dispersive tsunami model
BSD 2-Clause "Simplified" License
3 stars 0 forks source link

Where is the source code? #1

Open polxyxnn opened 4 months ago

polxyxnn commented 4 months ago

Hello,

I am interested in utilizing PComcot for research purposes. Is the source code still available/free/open to use?

More power!

Regards, Paul

yifanzhu-fluid commented 4 months ago

Hello, thank you for your interest in PCOMCOT. Don't worry, we will upload the source code and the manual very soon, which I think should be next month. A few months ago, we wrote a research article about PCOMCOT and submitted it to a journal. This article was accepted recently, so I will make all the source code open once our article is available online.

polxyxnn commented 4 months ago

Hello,

Thank you for the response. That is noted.

Congratulations on the success!

Regards, Paul

yifanzhu-fluid commented 2 months ago

Hi, Paul! I have uploaded the source code, examples, and a detailed manual of PCOMCOT. You can use it whenever you need. Just feel free to ask questions if you have problem using it. Regards, Yifan

polxyxnn commented 2 months ago

Hello Yifan,

Thank you for the update. Congratulations on the success! I will attempt to compile/install the code very soon. I will keep you posted.

Regards, Paul

polxyxnn commented 1 month ago

Hello Yifan,

I completely installed PCOMCOT in my PC and in a HPC environment. I attempted to run the modeling in these two platforms with 3 layers. For my PC, the run was successful but it was slower as expected. For the HPC environment, well, there was an error with Bi-CGSTAB algorithm:

ERROR: Bi-CGSTAB algorithm does not converge when solving dispersion term!

The said error was only present in the HPC environment, where I utilized 4 nodes with 16 processes allocated to each. PCOMCOT_1.txt Attached is the log file from the HPC environment for your reference.

Thank you so much!

Regards, Paul

yifanzhu-fluid commented 1 month ago

Hi Paul, I saw that the CFL number of your simulation is too large (2.184), which can cause the simulation to blast out quickly. You should use a smaller time step or coarser grids to ensure that the CFL number is about 0.5. Then, you can check if the results are stable with the Matlab scripts.

Regards, Yifan