Code of Conduct

Dmitri Naumov authored 5 years ago and Dmitri Naumov committed 5 years ago

It will not compile the default by reference capture
for some reason.

Thanks to @wenqing for reporting.

Fixed output times for iteration number based time stepping algorithm

Dmitri Naumov authored 5 years ago and Dmitri Naumov committed 5 years ago

Same as IterationNumberBasedTimeStepping.
The new function does not modify anything.

Dmitri Naumov authored 5 years ago and Dmitri Naumov committed 5 years ago

The new function does not modify anything.

Now this feature is used for the given time stepping
algorithm, the sequence of the output files has changed.

Keeping original references.

[MPL] Fix saturation dependent swelling.

Divide by dt to get a rate instead of an increment.
Update the test references.

Richards mechanics FEM fixes

[HM] enable mass lumping

IterationBasedTimeStepping fix FindMultiplier

[MPL] User-provided limits for porosity.

Dmitri Naumov authored 5 years ago and Dmitri Naumov committed 5 years ago

Currently the limits are hard coded to be 0 and 1.
In some cases these values could be higher and lower,
respectively.

Dmitri Naumov authored 5 years ago and Dmitri Naumov committed 5 years ago

Now, with the changing porosities etc. this might be
an important contribution.

Forces a multiplier < 1, when the Linear Solver
fails in an iteration step in it is > 1

Avoided the infinite time loop when a rejected step gives no change in the step size

Richards mechanics; Bishop's effective stress models

This was the assumption in the previous process' implementation.

Both models, power law and the saturation cut-off are used.

Now includes the Bishop's effective stress.