Low strain rate problems

[vitesh13]

At low strain rates, the coupling might happen too frequent. This takes the simulation in the current format, into small strain approximation.
To avoid this, we can adjust coupling frequency, but then by the time, there is enough displacement to warrant regridding, the time for SIBM is too high. This creates a microstructure with much lower dislocation densities.
So some small amount of SIBM is needed, which means that we need to CA at the intervals that you had thought correct, but without regridding

[vitesh13]

Steps:

• 1. Check how the normal restart works (does the stress strain curve simply continue?)

• 2. How will the orientation change work in normal restart - Mostly just need to change the rotational part of Fe (cannot do direct replacement without regridding), this means that Fp will correspondingly change.

• 3. How is displacement calculated after regridding is done? Does it get calculated based on the new geometry or it corresponds to older geometry?

If step 1 is good, that makes things a bit easier.
The most critical part is the step 2.

[vitesh13]

Test for point 2:
Do a DAMASK simulation and map to CA grid, then do two CA simulations:

1. Without any change in microstructure
2. One with some change in microstructure

Then load this geometry back into DAMASK and do 0 loading and see the orientations before and after reintroduction.