Page 1 of 1

New functionalities of MICRESS 7.0

Posted: Tue Jan 07, 2020 10:03 pm
by Bernd
Dear all,

A number of new functionalities have been included into Version 7.0:
  • As an alternative to the “classical” multiphase-field model, a reformulated version handles volume changes during phase transformations in an approximated manner. This is e.g. essential to simulate graphite growth in cast iron.
  • An extended restart-structure-only functionality (with zoom, shift and rotation) now also allows a change of dimensionality, e.g. to extract a 2D section from a 3D simulation. Note that the ‘structure-only’ input has been moved to the initial microstructure block. This block also contains the grain input part.
  • Temperature profiles over time can now be used also along with the moving frame option.
  • New option "ordered"/"disordered": These criteria can be specified in the "numerical parameters/concentration solver" section and can be used as alternative to "limits" or "criterion_lower/criterion_higher" in cases where an order-disorder transition is possible, like in γ/γ’ in Ni-based alloys. The criterion checks whether two sublattices are identical (disordered) or different (ordered).
  • The new functionality “split_from_grain” allows for the generation of separated parts of a grain, e.g. during melting. It is the analogue to the "add_to_grain" functionality and a reverse operation to "categorization". It comes as an option in the nucleation data and allows the user to address grain fragments which lost their spatial connection to the main part of the grain e.g. during melting. This function not only allows assigning a new grain number but also to change all the grain data as defined in the corresponding nucleation type (phase, orientation, shield data, etc.).
  • New facet, resp. faceted_c, model. The new approach combines the advantages of the two existing models faceted_a and faceted_b and will become standard in the future.
  • A new flag "penalty" replaces part of the functionality of the "limits" function in the "stoichiometric phases" input section. It provides a penalty term on the driving force for phase transformations and thus allows for avoiding negative or improper compositions under various circumstances. This new "penalty" function provides more control over the implied phase interaction and also strongly reduces undesired side-effects when using the penalty terms on the driving force.
  • The seed density distribution for nucleation can be specified as a log-normal distribution.
  • Stress Solver: A further homogenization approach for stresses and strains in the diffuse interface region- named Reuss/Sachs model - is now available which corresponds to the model explained in "I. Steinbach, M. Apel, Physica D 217 (2006) 153-160". All three elastic models available in the MICRESS “elastic module” (Voigt/Taylor, Khachaturyan, Reuss/Sachs) can now systematically be addressed for uncoupled (stress_u1, stress_u2, stress_u3), coupled (stress_ce1, stress_ce2, stress_ce3), and mechano-chemically coupled (stress_mc1, stress_mc2, stress_mc3) applications.
  • Stress Solver: For all three homogenization models (Voigt/Taylor, Khachaturyan, Reuss/Sachs), coupling to diffusion of elements can now be selected ("stress_mc1", "stress_mc2", "stress_mc3"). The model includes concentration dependent local eigenstrains, stress-driven diffusion, and stress-dependent quasi-equilibrium conditions.

Re: New functionalities of MICRESS 7.0

Posted: Wed Feb 19, 2020 5:32 pm
by Bernd
Hi all,

some more explanation on the new "penalty" option of MICRESS Version 7:

In MICRESS version 6.4, the "limits" option had 2 functions:

1.) Define upper and lower limits for the composition of an element in the specified phase, such narrowing the natural limits which are given by the database information on sublattices and their occupancies.
2.) Add a penalty term to the driving force in case the limit is exceeded. This penalty term has the function to promote interface movement in such a direction that the excess of the limit is reduced.

In MICRESS version 7.0, these two functions are separated:

1.) The option "limits" will change the natural composition limits of the phase as before, but no penalty term is added to the driving force. However, errors of the type 1xxyy with xx=phase, yy=element will appear, and recalculation of quasi-equilibrium will occur in case the limits are exceeded like in the older versions.
2.) The penalty term can now be explicitly added using "penalty <ph1> <ph2> <el>". The main advantages of this separation are that "penalty" can be omitted if it is not needed, and that "penalty" is narrowed down to a given phase pair only, while before it was applied to all interfaces of the limited phase to other phases. This narrowing strongly reduces the probability of side effects on the phase transformation kinetics.