[Orrin]

Hi,

I was wondering if anyone had encountered issues whilst trying to solve a sediment transport problem in REEF3D, using state files to enable benchmarking.

I used the I 40 and P 40 commands. The first simulation appeared to run fine and saved the state file folder. The second simulation successfully started running, but I noticed some issues with the 'Max Sediment Erosion' and 'Sediment Line' outputs.

The last sediment line reference for the first simulation is named '...._x-152770.dat'. This sediment line shows some sediment erosion as expected. The first 'sediment line' output from the second simulation is also labeled 152770 and shows the same erosion - which is as expected. However the following sediment line output from the second simulation (ref. 152780 in my case) returns to the initial sediment bed with zero erosion. This means that the second simulation does not appear to be using the previously benchmarked sediment bed, and is effectively just rerunning the original simulation.

Is there a known fix for this issue?

Thanks,
Orrin

[valgrinda]

Hi Orrin,

I will have a look and fix it for the next RC.

[valgrinda]

Hi Orrin,

I have tried restart for sediment. I cannot confirm your problem. For me the simulation continues as it should. I use the VTU files to check this.

[Orrin]

Originally Posted by valgrinda

Hi Orrin,

I have tried restart for sediment. I cannot confirm your problem. For me the simulation continues as it should. I use the VTU files to check this.

Hi Hans,

I only print the .VTU files at the end of my simulation (due to large file size), but the sediment line profiles suggest the sediment bed reset to the initial state after the first iteration.

I've just looked closely at the recent updates in GitHub and wonder if this could be due to the REEF3D version I am using. I was running V21.05 (for consistency of comparison with other simulations I've executed), and note in the update notes for V21.10 it states "x hot start support added (and working now) for sediment transport (CFD)".

Thanks,
Orrin

[valgrinda]

Hi Orrin,

yes, that is the problem. Please keep in mind to always use the newest version.