[MFWilliams]

Hello,


I am trying to run an FSI simulation and in my process to getting there I am trying to understand the solid mechanics part of it. However despite a lot of searching I have barely found any information on the solidProperties dictionary.


Please can somebody please explain the solid properties dicitonary and its entries as well as the different solid models available?



Also does the solid model chosen have any bearing on if you need to include pointD and pointDD in the 0 folder?


Thank you in advance
Merlin

[Hgholami]

Hello,
Your question is general. First you should look at the code, that where the solidProperties dictionary calles. then explain your question.

Originally Posted by MFWilliams

Hello,


I am trying to run an FSI simulation and in my process to getting there I am trying to understand the solid mechanics part of it. However despite a lot of searching I have barely found any information on the solidProperties dictionary.


Please can somebody please explain the solid properties dicitonary and its entries as well as the different solid models available?



Also does the solid model chosen have any bearing on if you need to include pointD and pointDD in the 0 folder?


Thank you in advance
Merlin

[MFWilliams]

Sorry for the general nature of the question it is because of my limited knowledge of the solidProperties dictionary and hence I am not exactly sure of the extent of what I do not know, if that makes sense.


1. I understand the solidModel prefix 'uns-' stands for unstructured but does this refer to the mesh being unstructured and so are these solidModels reccomended when using an unstructured mesh?


2. What is the difference between a 'totalLagrangian' model and an 'updatedLagrangian' model?



3. What is the difference between having the increment displacement or the total displacement as the primary unknown? Is the difference only noticed when post processing and analysing the results or must some simulations choose a specific one?



4. Do you determine which of D, DD, pointD and pointDD are required in 0 by reading the .C file in src/solids4FoamModels/solidModels?



5. What are the benefits of a block coupled solution over a segregated solution?


I hope that is slightly less general, and I would appreciate anyones time if they can answer them.


Best regards
Merlin

[Hgholami]

Hi

1. I understand the solidModel prefix 'uns-' stands for unstructured but does this refer to the mesh being unstructured and so are these solidModels reccomended when using an unstructured mesh?

For prefix "uns-", I recommend to compare the ::evolve() class of it with similar solid solver.

2. What is the difference between a 'totalLagrangian' model and an 'updatedLagrangian' model?

The different is in the theory of solve. look at the https://www.linkedin.com/pulse/total...linear-taneja/
for abstract: in updated Lagrangian, all variables being referred coressponding to the last calculated configuration. In total Lagrangian, all variables being referred coressponding to the initial configuration at time t=0.

3. What is the difference between having the increment displacement or the total displacement as the primary unknown? Is the difference only noticed when post processing and analysing the results or must some simulations choose a specific one?

It's again related to theory. All equation that uses in here should first read. some government equation derived base on assumption, for example small displacement, etc. So base on your problem, you should check the assumption and then choice the solver. otherwise, you should test each solver and validate it.

4. Do you determine which of D, DD, pointD and pointDD are required in 0 by reading the .C file in src/solids4FoamModels/solidModels?

The D, pointD is related to total Lagrangian, DD, pointDD is related to incremental. (DD is deltaD)
Trick: rename all D,DD,pointD,pointDD and run the each solid model. the solver give error and write each require file.

5. What are the benefits of a block coupled solution over a segregated solution?

The block coupled solution, make the large matrix to solve the variables but in segregated, the equation solve in segregated method. for detail look at
https://foam-extend.fsb.hr/wp-conten..._20Sep2018.pdf
In my expriment, the segregated method is faster than coupled in fluid solver. in solid solver, I don't experiment with coupled.