New Feature: Limits

new developments, future functionalities, hints for further applications of MICRESS
Post Reply
Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

New Feature: Limits

Post by Bernd » Sun Mar 29, 2009 1:39 pm

Sometimes, while using thermodynamic databases, problems can occur which are related to the composition sets of a phase. For example, in the TCFE6 database from Thermo-Calc MC carbides are modeled as FCC_A1#2, i.e., two phases, namely FCC and MC, are modeled as one common phase description. During a MICRESS simulation it may happen that for numerical reasons the carbon composition in a MC phase gets too low, and the phase suddenly "switches" to FCC. To prevent this, a new optional feature has been added, which can be specified in the input block where stoichiometric phases are defined:

...
Limits 2 7
40
50
...
no_more_stoichio

In this example, component 7 in phase 2 is limited to have a composition between 40 and 50 at%. Of course, the user is responsible for specifying reasonable values, the option should only be used if numerical problems occur and other means do not help.

Bernd

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

limits - keyword changed to lower case l

Post by Bernd » Thu Jan 25, 2018 5:38 pm

Hi all,

to comply to our standard, the keyword "limits" from version 6.4 on is completely in lower case. However, the old syntax "Limits" is still accepted.

Bernd

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Species in phase xxxx are not elements!

Post by Bernd » Thu Jan 25, 2018 5:59 pm

Hi,

sometimes during initialisation, MICRESS gives a warning like:

# Warning: Species in phase AL2O3 are not elements!
# Please use "limits" to manually set concentration limits!

This warning is not an error and can in many cases be ignored. When MICRESS initializes phases after reading phase data from database, a minimum and maximum concentration is automatically calculated based on the sublattice information from the database. This is done by analysing the minimum and maximum content of each element on each sublattice and summing up over the sites. However, there are cases where the database contains other species than elements. This is often the case in completely stoichiometric phases like Al2O3 where database developers prefered not to build up two sublattices (one for each element) but just define "Al2O3" as a single species o a single sublattice. MICRESS, in order to calculate maximum and minimum compositions for Al2O3, would need to analyse the stoichiometry of each species which can be very complicated and which is not done at present. Instead, MICRESS assumes standard minimum and maximum (0 and 1) and gives a warning.
In such a case, MICRESS will also not automatically detect whether the phase is stoichiometric for both elements Al and O. This then must be done manually to avoid problems. In case the elements are not completely stoichiometric, the user has the possibility to define composition limits manually by using the "limits" keyword.

Please note that using the "limits" keyword not only changes the minimum and maximum composition definition in MICRESS (which is used e.g. for correction of phase compositions before of after calling Thermo-Calc subroutines) but also activates a penalty on the driving force which helps to avoid the system from getting into these not allowed unphysical composition ranges. However, this mechanism can have unwanted side effects if the forbidden composition range cannot be avoided, especially if many limits are used simultaneously. The result can be wrong phase fractions or suddenly "jumping" interfaces. Therefore one should be careful and use "limits" only if really needed.

Bernd

Bernd
Posts: 1504
Joined: Mon Jun 23, 2008 9:29 pm

Limits: Error Code

Post by Bernd » Sun Apr 01, 2018 12:32 am

Hi all,

one thing which should be mentioned here is that the errors which are created due to user defined composition limits have a 5-digit number starting with 1:

"error = 10203"

would mean e.g. that in that a user-limit for phase 2 and component 3 was violated.

Bernd

Post Reply