[str6073]

I'm modelling a solar water heater with CFX. Solar radiation falls on tube portion and water circulates by natural convection . So initially my tank faces will be at ambient temperature. As the heating of water continues, temperature of faces also increases. How to give such a boundary condition ??? please help me.Its very urgent

[ghorrocks]

The increase in temperature of the face must be due to the thermal mass of the container. So if you want to model that thermal mass you need to model the material which forms the container. You need to place bounary conditions at locations where you can specify the thermal condition - so that might be adiabatic conditions on the outside of the container, or it could be convection to ambient air. These would both be valid options. But we do not know what your device looks like so cannot say whether it is appropriate for you or not.

[str6073]

Thanks for replying. I've attached my geometry figure with the thread. I've not considered the material of container.Only water is taken as the material.please look at the attached thumbnail.Its a simplified version of evacuated tube solar water heater

[ghorrocks]

If you are not considering the material then either an adiabatic or maybe convective boundary might make sense. As you have not explained much about your model I cannot tell which is better.

[str6073]

I've tried adiabatic condition.but the faces of the tank should experience a gradual temperature rise from ambient to the highest possible temperature.How can i give that condition?Keeping adiabatic conditon is sufficient ??

[ghorrocks]

Do you know what an adiabatic boundary condition means?

[str6073]

ya. There won't be any heat transfer to outside and from outside

[ghorrocks]

OK, good. So do you know what the heat transfer conditions are at the boundary? Is an adiabatic condition appropriate? If not, then what is the heat transfer condition?

You need to be able to know the heat transfer condition to specify a boundary condition - and the boundary condition you choose is one which models the heat transfer condition you know exists.

[str6073]

As i've considered only the fluid domain,faces of my tank represents water only.Geometry i'm considering is only a portion of the whole heater.So temperature on the faces of tank which i consider are influenced by temperature of water body near to them.In that way it can't be taken as adiabatic.Also by natural circulation,hot water will be on top portion and cold water will be at bottom.So temperature distribution over a face will be linearly increasing from bottom to top.And these faces will be at ambient temperatures at hte beginning of simulation.

[ghorrocks]

You appear to fundamentally misunderstand how CFD works.

Your fluid domain is where the fluid is, it is modelled by solving the flow equations. But the boundaries of that domain need to be specified with boundary conditions. These boundary conditions specify how the outer face of the domain interacts with the external world. So they are the interface from the domain to the outside world, not what happens inside the fluid domain.

So by adiabatic, I mean there is no heat flow from the fluid to the outside world. And on the domain side of the boundary there can be whatever flow the solver decides is suitable. And for natural convection boundary condition - I mean from the fluid domain to the outside world, not inside the fluid domain.

[str6073]

thank you bro. i really appreciate your help. keep in touch