Hallo MICRESS team,
first of all I wish you merry Christmas and a happy new year. Wish you also a big progress in further program development.
Today I have a question about separating the simulation into a few separate temperature profiles.
In my experiment, I quech down the austenite with high cooling rate to ferrite temperature. Then, hold at a constant temperature for gamma-alpha transformation. As my colleage suggests, lots of nucleation should have started since the quenching period. Then the ferrite grows from the settled nuclei during the isothermal transformation.
I still do not have a clue how I can make my simulation into 2 separated simulation interval: first interval-nucleation during high cooling rate + second continuous nucleation and/or growth of ferrite during isothermal transformation. Could you please suggest me?
Thank you very much in advance,
nokkikku
separated intervals of simulation
Re: separated intervals of simulation
Dear nokkikku,
wish you also a merry Christmas! Principially, there are two strategies how to treat different subsequent process stages in MICRESS: You can either implement them using only one driving file by defining several connection points in the thermal boundary condition input, or you can use several driving files and execute them subsequently by use of the restart option.
The first method has the advantage that you can run the complete process in only one job. Another important advantage is that if you need to change the numerical or physical parameter (e.g. for modification of the alloy composition) you need to do it in only one place. Thus, the method is less prone to a loss of overview over the input data.
On the other hand, especially for processes consisting of many steps, is may become more and more difficult to implement everything in one imput file, especially because many parameters cannot be changed apart from the possibility to define them temperature-dependent. Also, the total time may be confusing if the partial processes are taking place on different time-scales. Then, the restart option together with the optional keywore "reset_time" should be used (look here). This allows you to have each part of the process in one extra driving file which can be started using the restart output file from the previous part. Each part starts at the time t=0, you can change most of the parameters in the driving file like diffusion data, nucleation data, phase interaction data etc. between the different process parts (of course, you cannot change the size of the simulation domain or the number of elements!).
My advice: If you are just switching between cooling to holding, without changing temperature over too high ranges, physical and numerical parameters should be similar or identical in both process steps or only T-dependent. I would try to implement it in one single file if there is no obvious reason why this should not be possible.
Bernd
wish you also a merry Christmas! Principially, there are two strategies how to treat different subsequent process stages in MICRESS: You can either implement them using only one driving file by defining several connection points in the thermal boundary condition input, or you can use several driving files and execute them subsequently by use of the restart option.
The first method has the advantage that you can run the complete process in only one job. Another important advantage is that if you need to change the numerical or physical parameter (e.g. for modification of the alloy composition) you need to do it in only one place. Thus, the method is less prone to a loss of overview over the input data.
On the other hand, especially for processes consisting of many steps, is may become more and more difficult to implement everything in one imput file, especially because many parameters cannot be changed apart from the possibility to define them temperature-dependent. Also, the total time may be confusing if the partial processes are taking place on different time-scales. Then, the restart option together with the optional keywore "reset_time" should be used (look here). This allows you to have each part of the process in one extra driving file which can be started using the restart output file from the previous part. Each part starts at the time t=0, you can change most of the parameters in the driving file like diffusion data, nucleation data, phase interaction data etc. between the different process parts (of course, you cannot change the size of the simulation domain or the number of elements!).
My advice: If you are just switching between cooling to holding, without changing temperature over too high ranges, physical and numerical parameters should be similar or identical in both process steps or only T-dependent. I would try to implement it in one single file if there is no obvious reason why this should not be possible.
Bernd