[ajayjatoliya]

Hello REEF 3D Teams

I tried to solve the scouring problem due to the current only but getting velocity exceeding N61 frequently.
I have attached my ctrl file and control file, could you please look into them once
Thank you

[kamath]

You should check your results in Paraview, see where the velocities are too high. Then there is something to go on.

I also see that you define end of sediment layer at z=-0.3 m whereas your domain starts at 0. So there will be no sediment bed in the simulation

[ajayjatoliya]

Thank you, Kamath,
I tried to replicate the problem solved by Saud Afzal to calculate the scour around monopile due to current. he solved the problem for 72000 sec, but I don't know how to give input for simulation time which represents the actual time, is it model time (N 41) ?

[ajayjatoliya]

I want to model it for 72000 sec. so which time is representing it in simulations? I have attached a file could you please explain it

[Felix_Sp]

Hey,


I am no official member of the REEF3D team and am still learning. So beware, that my answers may be incorrect. I am trying to use the forum also as a learning platform.

In sediment transport simulations in REEF3D you have three sums of time. The computational time, which depends on your system ("total time" in the solver log, -> e.g. 70698.2 s in your screenshot). It consists of the time of all the mathematical operations took (e.g. "gctime") summed over the total simulation.

Then there is the solved time for fluid ("simtime" in the solver log). It is the time the fluid actually propagated though your domain (256 s in your screenshot).

And the sediment time (in your screenshot "Sediment Timestep ... etc."), which can be decoupled from the fluid time by the use of S15. With S 15 you decide whether the sediment transport time is identical with the fluid time (S 15 1) or whether the sediment is actually transported on a different timescale than the fluid (S 15 0 or S 15 2), thus decreasing the computational time as sediment transport is rather slow.

With S 15 0 (or S 15 2) you assume that the same fluid momentum acts on the sediment for a time calculated by S 13 and S 14. You can increase this time step by setting S 13 to a higher value. As fluid and bed are interacting with one another, such an assumption obviously becomes nonphysical at a certain value. Together with a high value of S 14 it may also become numerically unstable I guess.

Also, one wants a converged flow field before the sediment transport is actually initiated, thus the use of S 41 and S 42 (together with S 43 to S 46) is desirable I guess?

I think what you are interested in is the sediment time (in your screenshot "Sediment Timestep ... etc."), reaching a value of 72000 s. If you want to stop the sediment transport algorith there (as the printout may not match the 72000 s) you can use S 19 72000 to stop the sediment transport algorithm. Otherwise, you can use P 34 to print your results based on the sediment transport algorithm and not the fluid solver, which would also help in showing the correct time step.

Sorry for the long text, but I also want to use the sediment transport algorithms in the future and am actually interested whether Arun may correct anything I am saying, such that I also benefit from this discussion.

TL;DR I think what you are searching for is the "Sediment Timestep: ... etc." and in this line the "Total Time:" should reach the value 72000

Cheers
Felix

[kamath]

All correct Felix!

[Curiousli]

Originally Posted by Felix_Sp

Hey,


I am no official member of the REEF3D team and am still learning. So beware, that my answers may be incorrect. I am trying to use the forum also as a learning platform.

In sediment transport simulations in REEF3D you have three sums of time. The computational time, which depends on your system ("total time" in the solver log, -> e.g. 70698.2 s in your screenshot). It consists of the time of all the mathematical operations took (e.g. "gctime") summed over the total simulation.

Then there is the solved time for fluid ("simtime" in the solver log). It is the time the fluid actually propagated though your domain (256 s in your screenshot).

And the sediment time (in your screenshot "Sediment Timestep ... etc."), which can be decoupled from the fluid time by the use of S15. With S 15 you decide whether the sediment transport time is identical with the fluid time (S 15 1) or whether the sediment is actually transported on a different timescale than the fluid (S 15 0 or S 15 2), thus decreasing the computational time as sediment transport is rather slow.

With S 15 0 (or S 15 2) you assume that the same fluid momentum acts on the sediment for a time calculated by S 13 and S 14. You can increase this time step by setting S 13 to a higher value. As fluid and bed are interacting with one another, such an assumption obviously becomes nonphysical at a certain value. Together with a high value of S 14 it may also become numerically unstable I guess.

Also, one wants a converged flow field before the sediment transport is actually initiated, thus the use of S 41 and S 42 (together with S 43 to S 46) is desirable I guess?

I think what you are interested in is the sediment time (in your screenshot "Sediment Timestep ... etc."), reaching a value of 72000 s. If you want to stop the sediment transport algorith there (as the printout may not match the 72000 s) you can use S 19 72000 to stop the sediment transport algorithm. Otherwise, you can use P 34 to print your results based on the sediment transport algorithm and not the fluid solver, which would also help in showing the correct time step.

Sorry for the long text, but I also want to use the sediment transport algorithms in the future and am actually interested whether Arun may correct anything I am saying, such that I also benefit from this discussion.

TL;DR I think what you are searching for is the "Sediment Timestep: ... etc." and in this line the "Total Time:" should reach the value 72000

Cheers
Felix

Hello Felix,
Thank you for your wonderful response, which has cleared up a lot of questions I had! But I still have some confusion.
When I use S 15 0 or S 15 2, does it mean that the flow simulation time no longer corresponds to the sediment simulation time? It is when using S 15 1 that both time scales are 1:1?
For example, when using S 15 0, it shows in my Sedimentlog that by the time my Simtime reaches 18 S, the Sedtime has reached 1800 S. I would like to get the results of the monopile after 1800 S of flow scour, so should I stop the simulation at Sedtime=1800 S, or should I stop it at Simtime=1800 S ?
Thanks again for your great reply and I hope you can answer my doubts!

[Felix_Sp]

Hey there,
Yes you are right. Using S 15 0 or S 15 2 decouples the sediment time from the flow simulation time, assuming that the sediment transport time scales are much longer than the fluid transport time scales.
Depending on your settings (S 13, S 14), the shear force acting on the ground is assumed to act for the time specified in S13 and S14. The bed is moved and according to the flow it can react to the changed bathymetry/stream bed. If you want to achieve some equilibrium state, S13 and S14 will get you to your steady state faster.
Of course, if the flow can't react to the changed boundary with a sufficiently high frequency, your simulation can become unstable, or you have some kind of non-physical simulation.
Using S 15 1 with S 14 1 gives the motion of the boundary in the fluid flow timescale.
For example, if you analyze the temporal change under waves with a too large morphological time step and too high values for S 44/ S 46/ S 48. You may be missing/overestimating/underestimating the effect of the waves on the bed evolution.
TL;DR So if you want to analyze the erosion after x seconds, the sedimentation time is ALWAYS relevant.

[Curiousli]

Felix,thank you very much for your answer, it will help me a lot with my simulations afterward!