[Ehsan-Ar]

Hi there
I am modeling a large scale model of 30*10 kilometer and I faced problems in importing the original scale geometry in Design modeler even with large model support on. I wonder if it could be possible to scale down the original model, generate the mesh and again scale it up in Fluent. The main issue is about the effects of scaling up in fluent and if there are any limitations in Fluent, itself about a large model size.

Many thanks in advance.

[CeesH]

What kind of problems are you running into? Large model support, as far as I am aware, does not have to do with the physical size of the model but the level of detail (so, information size - bytes). Also for FLUENT itself, physiscally large domains should not be any issues (FLUENT is a dimensional solver, so there can be something as a "too large" model, but that would mean that our size somehow exceeds the largest possible size the memory can handle. Then we're talking about 10^308 meter, however).

Of course, when you work with very large models, there may be challenges capturing the relevant physics - you won't be able to accurately model boundary layers and bulk effects at these scales - but likely you are anyhow interested in one or the other, and if you can simple assume wall functions for the boundary layer, there is no reason why you could not model bulk flows at these scales.

SO you need to be a bit more specific about which issues you encounter, exactly.

[LuckyTran]

As mentioned Fluent is dimensional and writes natively in meters. I can't imagine what sort of problems you had that would go away if you scaled and unscaled it. The fact that you're using design modeler (which implies workbench) also makes this cumbersome. I've done 3 km models by accident. I would recommend to debug and figure out why you can't import your model.

[Ehsan-Ar]

Originally Posted by CeesH

What kind of problems are you running into? Large model support, as far as I am aware, does not have to do with the physical size of the model but the level of detail (so, information size - bytes). Also for FLUENT itself, physiscally large domains should not be any issues (FLUENT is a dimensional solver, so there can be something as a "too large" model, but that would mean that our size somehow exceeds the largest possible size the memory can handle. Then we're talking about 10^308 meter, however).

Of course, when you work with very large models, there may be challenges capturing the relevant physics - you won't be able to accurately model boundary layers and bulk effects at these scales - but likely you are anyhow interested in one or the other, and if you can simple assume wall functions for the boundary layer, there is no reason why you could not model bulk flows at these scales.

SO you need to be a bit more specific about which issues you encounter, exactly.

Thank you for your response
I am modeling a long dam(between 10 t0 30 kilometers) perpendicular to the shoreline and my objective is to simulate the water level difference between the left and right-hand side the dam as the tidal and ocean currents hit the structure of the dam. Consequently, adjacent cells have to be smaller due to the fact that the water level difference will be less than 2 or 3 meters. I am using VOF multiphase method for modeling the smaller scale.
As I mentioned, I created the large-scale geometry in Catia. The main Question is here: when I draw a large geometry in the Design modeler with large model support on, there isn't any problem with generating that model, but when it comes to importing my main Catia geometry into the design modeler, it can't generate imported geometries with dimensions larger than 1000 meters in any direction.
Thank you again for your reply

[CeesH]

well, if you want to simulate changes in the order of meters with a model of kilometers... good luck.
Is there any reason you need to model the entire dam? Can't you get away with a small section, e.g. few tens or hundred meters or so, or even a 2D cross section? What is the question that you are trying to answer and why does it need such a huge model to do so?

[Ehsan-Ar]

The main reason for the water level difference is the dam's length and shape, the longer the dam's length, the higher water level difference will achieve. As a result, I have to model the entire dam to see the effect of the length. As well as this, the level difference changed along the dam and I have to use a 3D model to observe the consequences. And about the question I am Answering, my main purpose is to see the effects of the dam's structure and shape on the ocean's streams near the dam.