Page 1 of 1

interface output

Posted: Tue Nov 24, 2009 3:43 pm
by ning
what is the exact formula used to calculate .intf output? Thanks.

Re: interface output

Posted: Tue Nov 24, 2009 4:27 pm
by Bernd
Dear ning,

The .intf is an integer output which is equal

1 in dual interfaces
3 in triple junctions
6 in quadruple junctions
10 in quintuple junctions
...

according to the number of combinations of pair-wise phase (grain) interactions. It is defined in all regions where the fractions are between phMin and 1-phMin plus the direct neighbour grid cells where one of the fractions is 0 or 1. In the remaining region, the value of .intf is 0.

Bernd

Re: interface output

Posted: Tue Nov 24, 2009 11:35 pm
by ning
However, EDX scan gives a smooth profile across the boundary...

Re: interface output

Posted: Wed Nov 25, 2009 12:47 pm
by Bernd
In the .intf output??????

I can imagine this only if n is very high so that you do not see the steps! But still...

Bernd

Re: interface output

Posted: Wed Nov 25, 2009 3:28 pm
by ning
Yes. intf scan gives a profile across bounday with 1 at the center and fading to 0 when merging to bulk.

That's why I thought that it should be related to phase variable profile across the boundary. You know, it is the key for calculation accuracy...

Thanks.

Re: interface output

Posted: Wed Nov 25, 2009 7:18 pm
by Bernd
Sorry,

I completely forgot that there is the extra output option "sharp" for the .intf file! If you activate this, the output is:

sum_over_pairwise_interactions ( 4 * frac_grain_i * frac_grain_j)

Then, for a dual interface, you have a maximum of 1 at the center of the interface and a continuous decay to both sides.


Bernd