Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Extension of Elmer in computational glaciology
Post Reply
alden
Posts: 11
Joined: 02 Jun 2020, 20:33
Antispam: Yes

Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by alden »

Hi all, I hope everyone is doing well and staying healthy.

I would like to add a thermally active bedrock body to a prognostic flowline model. If you start with the toy glacier example given in the elmer/ice courses is there a relatively straightforward way to do this?

The reason I ask is that I am interested in the thermal conditions at the bed of the ice, and it would be nice to have a more accurate representation of the physics.

I have done some searching of the documentation and haven't been able to find any obviously similar examples, and my own attempts have hit a wall against the "structured mesh mapper" solver.

Any thoughts would be great. Thanks!
raback
Site Admin
Posts: 4801
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by raback »

Hi

Do you need a temperature jump between bedrock and ice? If not it should not be too different. StructuredMeshMapper probably needs some mask to tell that it is active only at the ice.

-Peter
alden
Posts: 11
Joined: 02 Jun 2020, 20:33
Antispam: Yes

Re: Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by alden »

Hi,

I do not need a temperature jump. I will work on adding a mask to structured mesh mapper.

Thanks!
alden
Posts: 11
Joined: 02 Jun 2020, 20:33
Antispam: Yes

Re: Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by alden »

Hi, I spent some time trying to get to get a mask for the structured mesh mapper to work, using something like

Code: Select all

 Exported Variable i = string "material:ice"
but it was not really clear to me to what extent I needed to do this. As a result I kept running into the error

" StructuredMeshMapper: Sizes must agree: mesh velocity 2 "

So I figured I'd make an example sif and mesh starting from a toy glacier example to see if there may be some other problems with my approach.

My two main questions are:

Is my construction of the mesh and bodies correct?

How do I properly mask variables in this example so that structuredmeshmapper works?

Mesh Comments

To make the mesh I took the test glacier mesh, and extruded it into two layers. The first extrusion is the bedrock, and the second extrusion is the glacier. I mapped the surfaces and boundaries so that that the glacier extrusion corresponds to the original mapping in the toy glacier example. The bedrock layer was labeled body 3. The geo file needed to generate the mesh is attached below.

sif Comments

To keep everything simple, the third body added to the sif doesn't have any equations associated with it, and the rest of the simulation was not changed. Everything else works fine in this simulation if the structuedmeshmapper is not executed. It is only when I try to use this solver that the error above occurs. The sif does not contain any explicit masking of variables as described above The sif is attached below.

If you need any more information, let me know

Thanks!
toy_glacier_bedrock.sif
(6.73 KiB) Downloaded 448 times
testglacierbedrock.geo
(533 Bytes) Downloaded 443 times
raback
Site Admin
Posts: 4801
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by raback »

Hi

Look at test case "structmap6" and there the use of "Mapping Mask Variable".

This is treated in DetectExtrudedStructure of MeshUtils.F90

Have fun!

-Peter
raback
Site Admin
Posts: 4801
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by raback »

Actually "Mapping Mask Name" in test case structmap7 may be easier to use if you don't have a suitable variable living on the ice. This would create a permutation from the mask rather than using one from a existing variable. Of course if you have N-S then "Mapping Mask Variable" could be "Pressure" fox example.

Note that StructuredMeshMapper and StructuredProjectToPlane must both have the same masking.
alden
Posts: 11
Joined: 02 Jun 2020, 20:33
Antispam: Yes

Re: Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by alden »

Thank you so much! That worked like a charm. I used the "Mapping Mask Name" as in test case structmap7. I have attached the updated sif for anyone who is curious.

I appreciate your help!

toy_glacier_bedrock_with_mask.sif
(6.84 KiB) Downloaded 466 times
raback
Site Admin
Posts: 4801
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Adding a Thermally Active Bedrock Body to a Prognostic Flowline Model

Post by raback »

Hi Alden,

Great! Maybe you can share a picture too ;-)

-Peter
Post Reply