Hallo
I have a question about the diffusivity matrix in DICTRA/MICRESS. One can directly obtain the diffusivity matrix in DICTRA which is needed for the Fick's first law. For carbon flux in steel, for example, the concentration gradient of other alloying elements will affect the carbon diffusion flux. That's why we need to put other components in the diffusivity matrix (D fe,c,x) in the Fick's equation. Am I right?
I understand that the elements like Mo for sure reduces the diffusivity of carbon. If we have Fe-C-Mo system and if the Mo concentration is homogeneous, no concentration profile appears. How will the D fe,c,mo can affect the Jc? The term D fe,c,mo must be multiplied by the concentration gradient of Mo, which is now zero. Then the term relating Mo should be zero.
In this case, does only D fe,c,c affect the diffusion of carbon? To my understanding, it should not be like that because Mo should have at least some effect on carbon diffusion even the Mo concentration is homogenous. So D fe,c,c is anyway not the effective carbon diffusivity in the steel.
Can you give me any clue how one can obtain an 'effective' diffusivity of carbon under the influence of other alloying elements, or any reference?
Thank you very much and best regards,
nokkikku
Diffusivities matrix
Re: Diffusivities matrix
Dear nokkikku,
I have to admit that I am not an expert in DICTRA as you say, but the way the diffusion matrix is calculated in DICTRA should be the same as in MICRESS. Therefore, the problem would be the same if you would use MICRESS.
Essentially, if you have no Mo gradient, the corresponding off-diagonal fluxes have to be 0 (the complete column of the diffusion matrix) . But this does not mean that Mo has no influence on diffusion, because also the diagonal term for carbon depends on the thermodynamic factor (second derivative of the Gibbs energy) as well as on the intrinsic mobility which themselves can be depending on the Mo composition. But this should be well reflected in the diffusivity values which you get from DICTRA!
Hopefully, this answers your question above!
If you consider a moving interface, of course, the question would be whether you get a Mo gradient due to redistribution (ortho-equilibrium) or not (para-equilibrium). In this case, there would be an additional effect of the off-diagonal terms, because then you would have a local gradient close to the interface...
Bernd
Bernd
I have to admit that I am not an expert in DICTRA as you say, but the way the diffusion matrix is calculated in DICTRA should be the same as in MICRESS. Therefore, the problem would be the same if you would use MICRESS.
Essentially, if you have no Mo gradient, the corresponding off-diagonal fluxes have to be 0 (the complete column of the diffusion matrix) . But this does not mean that Mo has no influence on diffusion, because also the diagonal term for carbon depends on the thermodynamic factor (second derivative of the Gibbs energy) as well as on the intrinsic mobility which themselves can be depending on the Mo composition. But this should be well reflected in the diffusivity values which you get from DICTRA!
Hopefully, this answers your question above!
If you consider a moving interface, of course, the question would be whether you get a Mo gradient due to redistribution (ortho-equilibrium) or not (para-equilibrium). In this case, there would be an additional effect of the off-diagonal terms, because then you would have a local gradient close to the interface...
Bernd
Bernd