A question about material compositing (Multiple material models)

Hello all.

I’m trying to use a visco-plastic and latent heat model together.
Unexpectedly, I’ve faced a problem.

I’d like to select density from the latent heat model considering phase change and last materials from the visco-plastic model, but all parameters seem from the visco-plastic model.

I think the problem is that the visco-plastic model overwrites all parameters of the latent heat model even though I’ve chosen only viscosity for the visco-plastic model (it is shown below lines).

Any comment will be helpful for me :slight_smile:

Thanks

Best regards

Sungho

PS)

subsection Compositing
set Viscosity = visco plastic
set Density = latent heat
set Thermal expansion coefficient = latent heat
set Specific heat = latent heat
set Thermal conductivity = latent heat
set Compressibility = latent heat
set Entropy derivative pressure = latent heat
set Entropy derivative temperature = latent heat
set Reaction terms = latent heat
end

Hi Sungho,

Thanks for reporting your problem!
I just tried an example combining the Simple and Viscoplastic material models and I can confirm that for aspect at commit 617d400e0 Compositing is not working correctly, but it is for an older branch. There were some recent changes to the Compositing material model, so I’ll investigate further.

Cheers,
Anne

Update: this issue will be fixed by PR #3169. I think we should also add a test with or after that PR that combines 2 material models instead of only rerouting one through Compositing as is done in the current test.

1 Like

Hello Anne.

Thank you for your reply and you attention to my problem.
What is the meaning that this issue will be fixed by PR #3169?

Also, in your saying, combining two material models is done in the current test.
Does it mean parameters can be determined by two models like density are assigned by the multi-component model but also one of them is able to change phase by latent model?

Best

Sungho

Hi Sungho,

The development of ASPECT and its version control run through git and the GitHub platform (https://github.com/geodynamics/aspect). New contributions to the code are done in the form of pull requests (PRs) that go through some review process by the developers. What I meant was that there is an active PR on GitHub with number #3169 that fixes your problem. As soon as the PR is approved, you can check out and build the most recent version of ASPECT and get your model up and running.

ASPECT has a large test suite to ensure that PRs don’t break the current functionality. At the moment, we don’t have a test that covers your situation, and I merely stated that we should have such a test. When PR #3169 is approved and merged, you will be able to use the Compositing material model to have one material model determine e.g. the density and one e.g. the viscosity.

Cheers,
Anne

1 Like

Hello Anne.

Thank you for your explanation.
Your explanation helps me obviously understood the maintenance and development procedures of ASPECT in github’s system.

I am looking forward to PR #3169 will be approved as soon as possible :slight_smile:

Best

Sungho

I think PR #3169 is approved recently. I’ve downloaded and run the most recent version of ASPECT.
I can see the compositing model is working correctly.

But I can figure out something which I did not expect.

I choose visoco_plastic model and latent heat model by using composting model.
Latent heat model only for correcting density due to phase transition.
However, the model provide only reference density or another can be changed by transition even though I want to consider various compositions.

Could you give me any tip or any idea for how to consider viscoplastic model with multiple phase transition?

Hi Sungho,

We are currently working on a way to do multiple phase transitions in visco_plastic. This has been a while in the making, but is now a top priority and we hope to have this completed sometime in the fall.

There are two open pull requests related to this effort (#2656, #2776) and #2776 (creates general structure for handling phase transitions) should be merged soon. The big question then is how to make a general structure for handling multiple phase transitions and the properties (e.g., not just density) they may effect. One key issue is how should the properties for different materials and multiple phase transitions be handled in the parameter file. #2656 deals with this, but it may get reworked significantly based on recent discussions.

At this stage, it would actually be really helpful if you can provide a few examples of how you would like to use multiple phase transitions. Any and all feedback is welcome!

Cheers,
John