Page 2 of 2
Re: paraequilibrium simulation
Posted: Mon Mar 31, 2014 7:26 pm
by hchen
Dear Bernd:
it would be nice if you change your examples for the next release based on my case.
I have one more question about the kinetic transition in micress.
To my understanding, MICRESS in the "standard" mode should be possible to predict the transition from nple to ple if the cell size is small enough (smaller than the thickness of spike) . Am I right?
kind regards
Hao
Bernd wrote:Dear Hao,
Meanwhile, I analyzed the problem with "para" and found some conceptional issues which make it go "too fast", leading to the problem you described. In the next MICRESS version (6.2), the errors will be removed.
If you urgently need a fix, I can send you a corresponding executable. But the "paraTQ" and "nple" models are working correctly, so it should be fine for you...
By the way, your modifications inspired me to improve our standard example! With your permission, I will change our examples for the next release according to your case (with respect to the higher C and Mn compositions and to a constant temperature rather than a constant cooling rate)...
Best regards
Bernd
Re: paraequilibrium simulation
Posted: Mon Mar 31, 2014 8:04 pm
by Bernd
Hi Hao,
That is right!
But keep in mind how slow diffusion of substitutional elements is, and to simulate the nple-->ple transition, you also need to have the carbon peak fitting into the domain! Perhaps, it could be possible with a large 1D domain...
Bernd
Re: paraequilibrium simulation
Posted: Sat Apr 07, 2018 3:29 pm
by kamalnath
Hi Brend,
I have been looking in the forum to understand the different redistribution models used in MICRESS, epecially "para" and "nple". I presume "nple" will not allow for any partitioning but what does "para" do ? I tried looking up on "Paraequilibrium" but no use. And it is not described in the manual also.
Edit: And is it to possible add "anti-trapping" current in MICRESS ( Echebarria.B (2004) ) ? if yes, what is the option for it ?
Regards,
Kamal
Re: paraequilibrium simulation
Posted: Sat Apr 07, 2018 6:11 pm
by deepumaj1
Hi Kamal,
There are quite a lot of literature that explains the redistribution models NPLE and PE. I would like to quote a portion of one such literature here:
Different types of growth kinetics can be distinguished, depending on the redistribution behavior of the substitutional elements: paraequilibrium (PE), local equilibrium (LE), [12–14]and LE negligible partitioning (LENP). While for the PE process only the contribution of the fast diffusing interstitial elements is taken into account for the calculation of transformation kinetics, under LE conditions, the substitutional elements are also assumed to be redistributed; therefore, the transformation is alsocontrolled by the slow diffusion of the substitutional elements. Thus, transformations under LE may take many hours or even days, while under PE, they are finished in seconds. Both theories are limited by their assumptions and can be applied only in specific cases. LE is relevant for very small heating or cooling rates and PE for such processes as welding. LENP can be regarded as a special case of LE: if the driving force for phase transformation is high enough, the interface can move without being limited by diffusion of the substitutional elements, even if LE with full element redistribution on the atomic scale is assumed. As a consequence of the high transformation velocity, no partitioning is observed on the microscale. This is the most typical scenario for technical processes.
[Ref: J. RUDNIZKI, B. BO ̈TTGER, U. PRAHL, and W. BLECK. METALLURGICAL AND MATERIALS TRANSACTIONS A, 42A, 2516.]
Anti-trapping current is possible in micress. You can specify it under the 'phase diagram-input data' in the micress driving file. I'm not sure if this option is available in the 5.4 version that you use.
Deepu
Re: paraequilibrium simulation
Posted: Sat Apr 07, 2018 11:28 pm
by Bernd
Hi Kamal, hi Deepu,
I just want to add two things:
Firstly, in MICRESS 5.4, neither nple/para nor the antitrapping current have been available. Furthermore, those models have been improved in the course of the newer MICRESS versions, and I would recommend to use at least version 6.3. Otherwise, the models could work improperly under certain conditions. All this group of models is activated using the keyword "redistribution_control" in the phase interaction data input, and the individual models are defined for each element in the phase diagram input.
Secondly, antitrapping ("atc") cannot be combined with nple or para for the same element, and it would also not make any sense because the assumptions for these models are completely contrary. However, it would make very much sense to use e.g. "atc" for the fast-diffusing elements like carbon, and "nple" or "para" for the substitutional slow-diffusing elements.
Bernd
Re: paraequilibrium simulation
Posted: Sun Apr 08, 2018 4:00 am
by kamalnath
Thank you very much Deepu and Brend. I dont why I am not getting notifications for new posts. I have turned on the notifications from the user profile.
Deepu: As you have correctly predicted, I can't do ATC in my version of MICRESS
. But I found your reference very helpful.
Brend: Actually, v5.4 had "nple" and "para" option and it is working.
(
- image2.PNG (42.54 KiB) Viewed 4143 times
(look at line-364)
But as you have said, I will keep in mind that I may not get proper results. And as for as updating MICRESS is concerned, I will tell my advisor about the problems in old version.
Once again, thank you very much for your responses.
-Kamal