[lkortela]

Hello,

I am simulating the following using ANSYS CFX:

A kettle full of boiling water (100C) is placed into a sink full of
cold water (5C). I want to estimate how long it takes for the
boiling water to cool down to 10C.

Is it possible to simulate the following:

At time t, the cold water, which now has been heated up, is
replaced by cold water of temperature 5C?

For both fluid regions I defined the gravity. What value
should be used as the buoyancy reference temperature?

Cheers, Lauri

[JuPa]

This seems like a trivial problem surely you can estimate the time it takes by doing an energy and material balance.

Anyway, yes you can do this problem in CFX. Are you assuming there is no phase change (at 100 deg C are you assuming pure liquid, or saturated liquid/vapour)? If you're not assuming phase change then this would drastically reduce the complexity of your problem.

If you want to use CFX it sounds like will need to use the volume of fluid method, and the mixture model at the free surface to model splashing.

If you want to know how long it takes to cool down you'll need to do a transient simulation.

Your first port of call should be to complete the CFX tutorials. All of them.

P.S - you say buoyancy reference temperature - this assumes the Boussinesq approximation. The Boussinesq approximation is not valid for your temperature range (it is valid for small temperature differences - see the theory guide). I bet you'll get better results using variable IAPWS IF97 properties (instead of setting a ref temperature, it sets a ref density).

Edit: An alternative method might be to introduce a variable composition mixture of two materials in your domain, water cold and water hot. Set the initial condition of water cold to 5 deg C, and set the initial condition of water hot to 100 deg c. If this method works it should give you a very quick solution (compared to doing a full VOF method simulation).

[lkortela]

Originally Posted by RicochetJ

This seems like a trivial problem surely you can estimate the time it takes by doing an energy and material balance.

Anyway, yes you can do this problem in CFX. Are you assuming there is no phase change (at 100 deg C are you assuming pure liquid, or saturated liquid/vapour)? If you're not assuming phase change then this would drastically reduce the complexity of your problem.

If you want to use CFX it sounds like will need to use the volume of fluid method, and the mixture model at the free surface to model splashing.

If you want to know how long it takes to cool down you'll need to do a transient simulation.

Your first port of call should be to complete the CFX tutorials. All of them.

P.S - you say buoyancy reference temperature - this assumes the Boussinesq approximation. The Boussinesq approximation is not valid for your temperature range (it is valid for small temperature differences - see the theory guide). I bet you'll get better results using variable IAPWS IF97 properties (instead of setting a ref temperature, it sets a ref density).

Edit: An alternative method might be to introduce a variable composition mixture of two materials in your domain, water cold and water hot. Set the initial condition of water cold to 5 deg C, and set the initial condition of water hot to 100 deg c. If this method works it should give you a very quick solution (compared to doing a full VOF method simulation).

Thanks for the reply.

No phase changes.

I set material model for both domains to "water ideal gas" and defined the ref density. Had to switch to double precision to solve the model succesfully.

Do you have a take on how to model the changing of the cooling water? Maybe use results (temperature of the hot water and the kettle) as initial conditions in a new simulation and set again the cold water to 5C?

[ghorrocks]

Following up on Mr CFD's comment - this appears to be a trivial question to answer with a pen and paper approach. Why do you need to do it with CFD? What does it add?