Learning Forum

[toni.calabrese]

Dear ANSYS users,

I have a request for you and I hope someone can help me with that. I am new in the world of ANSYS. For a project a company asked us to perform simulations of a water storage placed in a lake. The water storage is submerged and anchored to the bottom of the lake. The storage is full of water with a temperature of approx. 80 °C and is insulated in order to keep the water inside the storage warm. The volume of the storage is not rigid because it is made of a flexible skin (something similar to PVC or similar not yet defined).

The goal of the project is:

1. evaluate the thermal losses of the storage
2. evaluate the forces, stress and strain of the system storage+anchorage
3. choose material of the storage skin and design anchorage that is able to bear the forces

Below you can see a sketch of the system to be modelled and simulated. The storage is affected by its weigth (Fw), the buoyancy force (Fb) and a force due to water streams in the lake (Fs).

Since I am new in ANSYS I am aware that I will need to learn the tool and probably use different tools within ANSYS.

My question is simple: which type of tools are needed in order to simulate such a system? I assume that firstly I would need a first simulation to find out thermal losses and forces (FLUENT?) and then perform a second simulation (Static Structural?) in order to test the materials and the anchorage under the forces found in the first simulation. Am I correct? Or could I perform some kind of coupled simulation to evaluate all these outputs?

Which kind of approach would you follow to reach the goal of the project?

Thanks and have a nice evening.

[Rob]

The appoach depends a little on what you expect to happen, and how precise an answer you want. 

Fluent would give you the heat transfer. Does the warm fluid cool, or is there a heating system? For the latter, that's steady state so a "simple" buoyancy problem. I'd be wary of modelling both warm and cold fluids simply for stability reasons. 

Mechanical would be the tool to determine the stress/strain/deformation. But... If you can assume the HTC values on the skin you could do it all in Mechanical and omit the Fluent model. 

[toni.calabrese]

Dear Rob, thanks a lot for your answer. I will try to be more precise in order to let you understand better which are our goal.

In a first step we will be interested on thermal performance of the storage. This means that we could test different geometry and calculate (with FLUENT?) the thermal losses of the storage. The goal here will be to check which geometry could minimize the heat transmission through the storage to the lake. There is any internal heating system in the storage. The storage should try to keep the water inside warm as much as possible since the storage is a seasonal thermal storage.

In a second step (once the geometry is fixed) our interest will be on materials and structure. Basically we would like to test (with the chosen structure) if the anchorage is able to absorbe the external forces and how the materials of the storage (that should be elastic) reacts on external forces.

Is now a little bit more clear? Thanks and I am waiting for your answer.

[Rob]

OK, thanks. Given the membrane insulation you may want to look at modelling just the warm side and use literature to find the external HTC for use in Fluent. There are too many factors to want to try and model the whole thing (hot and cold side) at once, not least baffles, external dimensions etc.  

Re the deformation. If you work out the volume at 80C and again at ambient temperature is there enough change in density to make a difference? 

[toni.calabrese]

Dear Rob, sorry but I am little bit confuse and I would like to proceed by steps. The first goal of the project is to evaluate the thermal performance of the system. Basically at the beginning we do not care about the structure and the stress on the materials. Only the thermal performance (under stationary conditions) is important to us. We would like to optimize the structure of the storage (shape, insulation layers and thickness and so on) and try to find an optized structure that minimizes the thermal losses.

If this is the first goal, which kind of tool should I use? Fluent would be enough?

[Rob]

If there's no change in shape then Fluent is the better tool. But you may be able to simplify the model by assuming a HTC to the lake rather than needing to model two buoyancy driven flows. 

[toni.calabrese]

Could you explain me better what you mean with the second sentence? If I assume the following:

1. storage does not move and does not change in shape;
2. constant Temperature of lake (10°C), storage (80°C) and air (5°C) (let's assume the storage is floating and the top is in contact with air)

With these assumption Fluent is not able to calculate the thermal losses of the storage in stationary status? 

[Rob]

Fluent can calculate all of the fluxes. However, the more "outside" you add to the model the larger the cell count, and you then need to model two sets of liquid buoyancy and now, the air space. If the storage is submerged you can neglect the air part. 

You may want to review the AIS Learning content to see what the various solvers do, and also how natural convection is modelled in Fluent. 

[toni.calabrese]

For practical reasons, it seems that the storage probably should be placed on the surface of the lake. This means that the top of the storage will be in contact with the air and the walls and bottom will be within the lake. Based on your suggestion the simplest approach will be to model only the storage and the volume of water around the storage. In order to take into account also the thermal losses through the top, your idea would be to set a constant temperature for the air but avoid to have also a domain that represetns the air above the storage. Right?

[Rob]

It would be easier to not need to deal with the air as well. You're then making more assumptions (wind over water causes motion) or using yet another model in Fluent that's likely overkill for what you need. A big part of using simulation is understanding the problem before going anywhere near a computer. Additionally, my ability to offer advice on here is limited as it's classed as "public" for the purposes of Export Law. 

[toni.calabrese]

OK, thanks for the suggestions nad I wish you a nice week end