Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Jerry1999
    @Jerry1999
    opencalphad/opencalphad is now open.
    Bo Sundman
    @sundmanbo
    This room is open for questions and discussion about Computational Thermodynamics in general and OpenCalphad in particular
    Marat Latypov
    @latmarat
    Hi, could you please advise where to find input files for uphill diffusion example ?
    Bo Sundman
    @sundmanbo
    I assume you mean the AlNiPt example. All files for this are on the directoryexamples/TQ4lib/F90/parallel-alnipt/ directory. There is a README.txt explaining how to compile and link the example program. The actual input to the program is the file "setup.dat"
    2 replies
    Bo Sundman
    @sundmanbo
    @latmarat @sundmanbo The alnipt example provided with the OC software is also an uphill diffusion.
    VictHe
    @VictHe
    Hello everyone, I have no chemical background, therefore might have a rather simple question about computational thermodynamics: How does OpenCalphad calculate the chemical activity of a component in a phase? Does the chemical activity depend on whether there is a phase change or not? If so, how? Thanks a lot in advance!
    Bo Sundman
    @sundmanbo
    We wrote a paper last year in the Calphad journal (https://doi.org/10.1016/j.calphad.2021.102330) explaining how OC calculates equilibria as well as diagrams. You can find other references in that paper. A quick answer is OC users Lagrange multipliers to handle external conditions as well as internal constraints.
    petrvokac
    @petrvokac
    Can OpenCalphad read "dat" format instead of "tdb"? For example https://ars.els-cdn.com/content/image/1-s2.0-S0364591613000308-mmc1.zip
    Bo Sundman
    @sundmanbo
    No. There are some DAT to TDB converters but I do not know where to find them. There are several free databases in TDB format available, some have links from the OC homepage. Which systems are you interested in?
    petrvokac
    @petrvokac
    Thank you. I started to follow literature and I am collecting databases provided as open supplementary data. Just to test them and learn what is available. Now I am investigating Fe-O system. I found feo_kje_sun.tdb which works with just with few modifications. Then some other oxide systems with a goal to learn how to plot Ellingham diagrams.
    Bo Sundman
    @sundmanbo
    Ellingham diagrams are not really useful, they consider only stoichiometric oxides wheras in reality all oxides (as well as metals) are solution phases. OC cannot calculate Ellingham diagram except one phase at a time which is a bit tedious. But if you include all oxides at the same time you can calculate which are stable for a given alloy composition at varying T and oxygen potential. My experience is that many of the traditional tools for corrosion or other applications become redundant using computational thermodynamics but one has to look at the problem from a different perspective. You do not have to check for each metal, you can specify your alloy composition directly.
    Bo Sundman
    @sundmanbo
    I have a database with Al-Fe-Si-Ni-Ti-C-O (in TDB format using the ionic liquid model) and a few more elements which is some 20 years old but still reasonable I think. If you are interested you can contact me directly.
    ksvtm
    @ksvtm:matrix.org
    [m]
    Hello there. I"m trying to build pyOC from recent commits using autotools. after running ./configure --with-python and make it seems that it produces libraries libOPENCALPHAD and libpyOC that are not linked correctly, running nm on that libraries shows that links to metlib and liboctq are missing. Can't really understand how to fix autotools scripts for that, because i can't understand what part of configure script produces libOPENCALPHAD and libpyOC. I'd be grateful if someone helps me with that issue. Thank you.
    2 replies
    zymnus
    @zymnus
    Hello everyone, I wanna ask a problem about the output from OC software. I use the OC6A to calculate the enthalpy of NiTi system. Actually I can successfully get the T-Hm plot for x(Ti)=0.2, but I don't know how to output the calculated T-Hm data as .dat or other kind of data file from OC6. Is it possible to do it in OC6? Thanks in advance!
    5 replies
    pooja19gangare
    @pooja19gangare
    Hello OC users. I want to know that how to compile the OC?
    3 replies
    pooja19gangare
    @pooja19gangare
    Thank you for the suggestion. I have another query. That is i want to perform the phase diagram calculation for my build system. How to do it? I mean how to create the TDB format ?
    5 replies
    pooja19gangare
    @pooja19gangare
    Thank you for the reply @sundmanbo. The build system refers the system constructed by myself. I'm working on Ni Based alloys, in which I'm looking into the grain boundaries. So far, i have determined the various properties related those grain boundaries using VASP. But I need to know the phase diagram of the systems generated/constructed by me (that is Ni Based Alloys).
    It is possible to do it?
    Bo Sundman
    @sundmanbo
    OC deals only with bulk properties and often grain boundary energies are estimated from these but of course one can also estimate them by DFT. But the phase diagram is independent of the grain boundary energies (all interfaces are ignored). Thus the TDB file does not contain such values. They have to be added by the simulation software you use for grain growth or phase transformation. This must also take the geometry and curvature of the boundary in consideration, preferably 3D. There is a free (CC BY license) carefully collected HEA database I mentioned in another reply. I can send it if you contact me by email.
    Bo Sundman
    @sundmanbo
    As answer to a question asked by email the PURE database has only data for the pure elements. It not possible to calculate any phase diagram using this database. Assessment of binary and ternary systems is a big effort needed to develop databases to obtain phase diagram that correspond to the experimental data. All of you who benefit from the work already done developing the Calphad databases which are available today should think about contributing to extending these by devoting some efforts to learn to assess some systems.
    Sakimi Yamazaki
    @Sakimi_Yamazaki_twitter
    Nice to meet you. I have a question: how can I use OCASI to calculate a single -phase free energy and its concentration derivative?
    The subroutine tqgetv seems to support this, but I don't know how to use it. I would be very grateful if you could tell me how to use it.
    1 reply
    Sakimi Yamazaki
    @Sakimi_Yamazaki_twitter
    Thank you for your prompt reply. I have been writing code in fortran for about 20 years.
    Bo Sundman
    @sundmanbo
    Nice, It is rare to find someone programming in Fortran today. I assume you can use the example I provide in examples/TQ4lib/F90/parallel-alnipt/ to create a program to set conditions to calculate an equilibrium (the crfe and feni examples are not really up-to-date). If you then want to extract values of state functions and variables such as Gibbs energy, enthalpy, chemical potentials etc. you can use the subroutine "tqgetv". However, to extract partial derivatives of the Gibbs energy with respect to constituent fractions, "dG/dy_i", you must use directly variables in the data structure TYPE(gtp_phase_varres). For example %dgval(1,i,1) is the partial derivative of G with respect to constituent "i" (before the % you must specify the phase and I assume you can find examples in my code how to specify this). Note that "i" is a constituent variable index in the phase, if you want partial derivatives with respect to components or elements you may have to combine several such partial derivatives. You find the definition of all TYPE structures in the src/models/ directory or in the doc/models/gtp.pdf documentation. I have recently updated this on the github page. Please read some of the documentation how to obtain for example the chemical potentials for an endmember "I" using partial derivatives fore a phase with a CEF model the equation is mu_I = G_M + \sum_s (dG_M/dy_s,i)_TPy_k - sum_s\sum_j y_s,j (dG_M/dy_s,j)_TPy_k ..... using just a single partial derivatives is not a physical property.
    Sakimi Yamazaki
    @Sakimi_Yamazaki_twitter
    Thank you again for your kind reply. I will try to read the materials you told me and try to get G as a function of composition.
    If it is difficult, I may write a code, reading my TDB file and calculating G and Gc by myself. I do not have the skills to do equilibrium calculations, so I will make use of your code. I will let you know if I complete my desired calculation program.
    2 replies
    Marat Latypov
    @latmarat
    Dear Dr. Sundman (@sundmanbo), I am currently learning the assessment methodology in CALPHAD. I can see that the assessment depends on a researcher's decisions on which literature data to include.
    I am wondering if you know some examples of simple (e.g. binary) yet "controversial" systems, in which calculated phase diagrams from two (or more) assessments differ significantly because of difference in researcher's included data.
    Thank you!
    2 replies
    Bo Sundman
    @sundmanbo
    There is a command LIST EXCELL_CSV to list the results from a STEP command in a CSV table.
    markmikers
    @markmikers
    Hello. I am completely new to OC/CALPHAD. For multi-component systems, is it possible to vary a composition of multiple elements at the same time in STEP or MAP calculations? E.g., to create a 2-dimensional pseudo-binary x-T phase diagram for Fe-Cr-Ni where ends of x axis are Fe3Ni and Cr?
    Bo Sundman
    @sundmanbo
    image.png
    The attached diagram for Fe3Ni-Cr. Some lines are not complete, with OC one often has to use several start points to get the full diagram
    3 replies
    Bo Sundman
    @sundmanbo
    You cannot set arbitrary start and end compositions but the section you suggest can be obtained by the conditions T=1000 p=1e5 n=1 x(fe)-3*x(ni)=0 x(cr)=.3. Then set axis 1 to x(cr) and axis 2 to T. In OC you can use linear expressions of state variables of the same type as conditions. I recommend you to look at the OC examples to find other ways of setting conditions.
    Bo Sundman
    @sundmanbo
    image.png
    Bo Sundman
    @sundmanbo
    An iso-pleth diagram can be difficult to understand unless calculated in several different ways. In the iso-pleth (I do not like the term pseudo-) there seems to be a thin region with a single sigma phase around x(cr)=0.55. This also shows up in the STEP plot above calculated using T=1100K. Below 800K the diagram is complicated. I have not considered the ordering in the FCC phase at the Fe-Ni side.
    Bo Sundman
    @sundmanbo
    Finally the isothermal section at 1100K. The iso-pleth is the dashed line cutting the top of the sigma phase. Note also the Ni-rich FCC comes back at the high Cr content
    Bo Sundman
    @sundmanbo
    image.png
    markmikers
    @markmikers
    Dear @sundmanbo, thank you very much for a detailed response. I am now trying to make similar plots for a system with more elements and it seems working.
    Pranjali Khajanji
    @khajanji:matrix.org
    [m]
    Can I please know if we have any database for polymeric systems? Or something related to battery research?
    Bo Sundman
    @sundmanbo
    I only provide the OC software with some small databases needed for testing. The UNIQUAC model for polymers is implemented in OC but you have to search for databases elsewhere as well for battery systems.
    Aung1234Auto
    @Aung1234Auto
    Hello, Opencalphad community! I have been using Opencalphad for my materials. I have tried to predict the phase equilibria of the materials with more than 4 components. Actually, I was able to plot phase equilibria and property diagrams. But, I would like to understand the mathematical expression used in the calculation of this multi-component system. Thanks!
    Bo Sundman
    @sundmanbo
    It is not really a "mathematical expression", rather a sequence of algorithms. I suggest you start by reading the paper by Dupin, Hallstedt and me in the Calphad https://doi.org/10.1016/j.calphad.2021.102330. There are references there to learn more.
    1 reply