Improve EVP rheology

[simone-silvestri]

This PR improves a bit the interface of the EVP rheology by including

• a parameter that controls the value of the relaxation field
• a parameter that allows the choice of the ice pressure between a simple ice strength and a replacement pressure (which is the default and only option in main but seems like it takes more to converge)

[glwagner]

maybe better to introduce a property called pressure_formulation. This is more extensible (eg more than two options), and avoids the complexity of a floating type parameter

[simone-silvestri]

Apparently, @jm-c said that the replacement pressure (which is what we have now by default) might be the cause of the big pile-up that occurs using EVP near immersed boundaries.
This is because when $\Delta \approx \Delta_{min}$ then we are artificially reducing the ice strength following
$\tilde{P} = \frac{P \Delta}{\Delta + \Delta_{min}}$, so it might be that (when velocities are low near the immersed boundaries) we are pushing towards the immersed boundaries and the ice gets stuck there.

Maybe changing the formulation to $\tilde{P} = \frac{P \Delta}{max(\Delta, \Delta_{min})}$ might work?

[glwagner]

Apparently, @jm-c said that the replacement pressure (which is what we have now by default) might be the cause of the big pile-up that occurs using EVP near immersed boundaries. This is because when Δ ≈ Δ m i n then we are artificially reducing the ice strength following P ~ = P Δ Δ + Δ m i n , so it might be that (when velocities are low near the immersed boundaries) we are pushing towards the immersed boundaries and the ice gets stuck there.

Maybe changing the formulation to P ~ = P Δ m a x ( Δ , Δ m i n ) might work?

Seems like a good thing to try, you can add the alternatives as new formulations

[glwagner]

Is this the case with the max in the denominator? Is there a way to have a more descriptive name?

[simone-silvestri]

We could put the operation of the denominator as a field of the ReplacementPressure (max or +) so that the formula would be return P * Δ / operation(Δ, Δmin)

[simone-silvestri]

I removed ModifiedPressureReplacement for the moment to merge this, we can keep it in account and add it later when we want to test new formulations.

[codecov]

Codecov Report

Attention: Patch coverage is 92.30769% with 1 line in your changes missing coverage. Please review.

Project coverage is 70.44%. Comparing base (f22a998) to head (4e606dc).
Report is 5 commits behind head on main.

Files with missing lines	Patch %	Lines
src/Rheologies/elasto_visco_plastic_rheology.jl	92.30%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #66      +/-   ##
==========================================
- Coverage   70.71%   70.44%   -0.27%     
==========================================
  Files          27       27              
  Lines         816      829      +13     
==========================================
+ Hits          577      584       +7     
- Misses        239      245       +6     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.