Hello,
I came across some problems when I tried to do some MICRESS simulations about alloy solidifications:
1) Is there a way to plot the temperature-distance profile at some time by using some file generated during MICRESS simulation?
2) We usually see the tip cooling rates, and tip temperature gradients in published papers about alloys solidification, but we need to input the cooling rate and temperature gradient in .dri file. Are there some relations between those tip ones and those input in MICRESS .dri files?
3) We usually input constant diffusivities or temperature-dependent diffusivities in .dri file if we do not couple an atomic mobility database. Is there also a way to enter the composition-dependent diffusivities in MICRESS? It should be helpful when we study diffusion cases at constant temperature.
Thank you in advance for your help!
Some quesitons about solidification simulaitons
Re: Some quesitons about solidification simulaitons
Hi sunny,
welcome to the MICRESS forum! I will try to answer your questions:
1.) For plotting temperature-distance profiles you can use the .temp output and the DP_MICRESS postprocessor. In order to get the .temp output, you need to specify out_temp in the outputs selection. For alloy simulation (i.e. in the case of concentration coupling) the temperature is changing only in z direction, so temperature-distance profiles should be plotted parallel to the z axis.
2.) In the standard case, one can assume a linear temperature profile on the microscopic scale. This is the reason, why you are specifying a cooling rate and a temperature gradient. Then, of course, the assumption is that the temperature gradient at the dendrite tip cannot change, only the temperature can change. The relation between the temperature at the tip and at the bottom (where the input is referring to) depends only on the distance bottom-tip and the given temperature gradient.
If you think this approximation is too crude, then you can use the 1d_temp option which applies an external one-dimensional temperature field. But this implies that you also have to specify the thermal boundary conditions (look here).
3.) It is not possible to "manually" input a concentration dependent diffusion coefficient. We assume that normally the user has no information about concentration dependency if not from thermodynamic or mobility databases.
Anyway, the main contribution typically comes not from the concentration dependency of the mobility, but of the thermodynamic factor. So, if you do not have a mobility database for your system, but a thermodynamic one, then one option would be to create a "dummy" mobility database with constant mobilities. Then the concentration dependency due to the thermodynamic factor would be automatically included.
Best regards
Bernd
welcome to the MICRESS forum! I will try to answer your questions:
1.) For plotting temperature-distance profiles you can use the .temp output and the DP_MICRESS postprocessor. In order to get the .temp output, you need to specify out_temp in the outputs selection. For alloy simulation (i.e. in the case of concentration coupling) the temperature is changing only in z direction, so temperature-distance profiles should be plotted parallel to the z axis.
2.) In the standard case, one can assume a linear temperature profile on the microscopic scale. This is the reason, why you are specifying a cooling rate and a temperature gradient. Then, of course, the assumption is that the temperature gradient at the dendrite tip cannot change, only the temperature can change. The relation between the temperature at the tip and at the bottom (where the input is referring to) depends only on the distance bottom-tip and the given temperature gradient.
If you think this approximation is too crude, then you can use the 1d_temp option which applies an external one-dimensional temperature field. But this implies that you also have to specify the thermal boundary conditions (look here).
3.) It is not possible to "manually" input a concentration dependent diffusion coefficient. We assume that normally the user has no information about concentration dependency if not from thermodynamic or mobility databases.
Anyway, the main contribution typically comes not from the concentration dependency of the mobility, but of the thermodynamic factor. So, if you do not have a mobility database for your system, but a thermodynamic one, then one option would be to create a "dummy" mobility database with constant mobilities. Then the concentration dependency due to the thermodynamic factor would be automatically included.
Best regards
Bernd
Re: Some quesitons about solidification simulaitons
Hi, Bernd,
Thank you very much for your explanation! I think I know more about it now.
And I also interested in your mentioned "1d_temp" option in MICRESS, and I wonder whether this option can only used in the case of concentration-coupled and latent heat considering?
Thank you!
sunny
Thank you very much for your explanation! I think I know more about it now.
And I also interested in your mentioned "1d_temp" option in MICRESS, and I wonder whether this option can only used in the case of concentration-coupled and latent heat considering?
Thank you!
sunny
Re: Some quesitons about solidification simulaitons
Dear sunny,
that is right! The 1d_temp option is meant as an alternative to the pure temperature coupling,i.e. for the case of alloys. As temperature is explicitely solved in the 1d temperature field, using latent heat is obligatory. I just finished a beta version which treats this more clearly and which removes some bugs which occured when trying to use 1d_temp for grain growth. It will be available in the next days.
Bernd
that is right! The 1d_temp option is meant as an alternative to the pure temperature coupling,i.e. for the case of alloys. As temperature is explicitely solved in the 1d temperature field, using latent heat is obligatory. I just finished a beta version which treats this more clearly and which removes some bugs which occured when trying to use 1d_temp for grain growth. It will be available in the next days.
Bernd