>One comment regarding ARIT. I didn't go carefully through all the results
>but my impression is your values are a bit underestimated. If you added
>additional Gaussian term to your An, you should get results much closer to
>WA. That should be relatively simple in your model (I think:-)? Of course,
>that is probably an oversimplification again, but it would get you much
>closer to the real thing?
I did not try to apply WA, too boring ;-). In such a case, the best
WA result would be obtained by using all the reflections, not only
2 harmonics, otherwise, the WA method can lead to quite
underestimated results too. So I wait for the official results in
order to see how underestimated my results are...
About adding a Gaussian term to the An, that is not as simple
as you think. The strain model in ARIT contains already the
possibility to correspond to a large range of profile shapes,
including Gaussian. But the size model is currently only
Lorentzian in ARIT. There is no way to have a pure Gaussian
shape for the size effect since it corresponds to unphysical
size distribution function (negative proportions of
cell columns for some defined lengths...). But if you
have a versatile size distribution function which would
correspond to a simple analytical AnS (size only) Fourier
series, with few parameters, please tell me about it. It would
be probably easy to include inside ARIT. Nevertheless, I do not
believe in a simple solution to that problem. Several size
distribution models could be tested too, unimodal, bimodal,
multimodal, Gaussian (for the size distribution, not for the
profile shape), Poisson, etc. And you have to consider the
general case : anisotropic, contrarily to the CeO2 sample.
So that, the size distribution functions will be possibly
different, according to the crystallographic direction.
Mathematics behind are not simple, that would ensure
a globally coherent model. This explains why ARIT stays
at its 1984 development level.
Best,
Armel Le Bail
http://sdpd.univ-lemans.fr/course/
