Issues combining latent heat model with viscosity structure

ASPECT
1

Hi there,

  I am modeling a 3D convecting system, and trying some model ingredients. I am specifically trying to incorporate phase changes, just one for now, at a certain depth. This is easily done with the latent heat model..specifying Clapeyron slope, density jump, transition depth and width, etc.
 I am finding trouble when trying to combine that with an ambient viscosity structure and a compositional field. I am using the Depth dependent model, providing a viscosity layering. As a base model I am using the latent heat model.

The model does not run. The core of the error message is this: ''Exception 'ExcMessage("Error: At least one list that gives input parameters for the phase " "transitions has the wrong size. Currently checking against transition depths''

This seems easy to fix by writing and specifying the correct number of viscosity layers, and the correct number of phase transition interfaces.

In my model there is only one phase transition (for now) and several viscosity layers, and one compositional field.

I have been unable to find the right combination of specified interfaces of each kind. I have tried yes/no including the phase transition as a viscosity interface.

Do I need to, in addition of providing viscosity and depth lists, write the viscosity layers inside the latent heat model to ghost-disguise them as zero-Clapeyron-slope interfaces?

I would appreciate suggestions from someone who has tried this kind of combinations.

cheers many !
Felipe

2

Hi Felipe,

sorry for the late reply.
The error message you get from the latent heat model (At least one list that gives input parameters for the phase transitions has the wrong size) only corresponds to the parameters you use in the latent heat material model (so they are independent of what you do in your depth-dependent viscosity model), but they need to be internally consistent. So if you have one phase transition, you need one entry in “Phase transition density jumps”, “Phase transition Clapeyron slopes”, etc; and two entries in “Viscosity prefactors”. If you look at the documentation of the parameters, all the ones that have a specific requirement for how many entries they need to have should say that in their description.

So, to answer your question, yes, you need to have two entries in the “Viscosity prefactors” parameter (and these two entries can be have same value). You can have as many viscosity jumps in your depth-dependent model as you want, but those will always have a zero Clapeyron slope (that’s just the setup of the model, everything only depends on depth).

If you want to look at an example for an input file with one phase transition using the latent heat model, you can take a look at the latent-heat.prm cookbook.

Best regards,
Juliane