Learning Forum

[Phuc Hai Dang Nguyen]

Hello, 

I'm currently trying to figure out if I can simulate the sediment (for example sand) transport at an artificial beach with ANSYS fluent. For my thesis I have to build a modell with waves/water and I have to analyse how the water affects the sediment under it. The purpose behind it would be, if that artifical beach could be used for research purposes in real life or if sand would just be piling up at one place.

I know that two phases are possible, however it was hard for me to search for the right information. Would you also have links or similar projects, where I could see how something like that is build? The internet doesn't show anything specific to my problem.

I'm thanking you for your help.

Kind regards

[Rob]

Not easily. The wave model is only available with VOF (for now). Granular flows need Mixture or Eulerian in Fluent, or Ansys Rocky. The latter may be overkill and need too many particles to be feasible. 

The next problem is where the sand comes from and goes to. If you look at coastal erosion sand may be moved several miles and the cause of currents/waves/channels may be due to something much further up the coast. Modelling that with CFD would be challenging. 

For localised stuff (ie piers, break waters etc) then CFD can be used (we've done it) for larger systems there are 2.5D codes such as TUFLOW.  

[Phuc Hai Dang Nguyen]

Thank you very much for your reply.

The model would be for an hypothetical surf parc where artificial waves can be made and the question would be if sand could be used for the aesthetics but also if this sand within the surf park can be used for research purposes. To my knowledge, the sand would be the the floor bed on top of the surf parc. 

Does this give a bit more insight on my problem? Would you still say that it would be too difficult to build 

[Rob]

OK, so the system is a limited domain, ie a large wave tank. It's doable, and if you create waves using a paddle system (like in a wave tank) you can use multifluid VOF (Eulerian) so can include a granular phase. It's not going to be a quick, small or simple model though. 

[Phuc Hai Dang Nguyen]

Thank you for your very helpful advice! Do you perhaps have any specific links to the projects? Or are they availbe over the Ansys learning platform? As it is a model for a master thesis do you think, it would be doable in a frame of 3-4 months?

[Rob]

I don't have any links, most of what we've done was for a specific customer so can't be shared. 

The wave tank part is doable for 3-4 month project assuming you're reasonably proficient with the solver and have a decent computer. Adding the granular part isn't trivial, so I'd be more wary of that. Difficult to say for sure as I don't know your skill level or how much supervisor support you'd get.