Re: Nonlinear constraints in RIETAN?

Fujio IZUMI ( izumi@nirim.go.jp )
Thu, 15 Feb 1996 13:22:53 +0900

Dear Dr. Estifanos and other users of RIETAN-94:

>I would like to include nonlinear constraints (ca. 20) in RIETAN
>(rietanppc). First an output file of ORFFE (*.ffe) must be created. Using
>the sample input file fapatite.ins, I could calculate interatomic distances
>but for the bond angles there is an error message
>"BOND ANGLE IN DEGREES. CENTRAL ATOM IS VERTEX
>
>### Run time error - Premature end of file" .
>How should I include the instructions in the *.ins file (If NDA > 0 then
>...} end if) or in the *.xyz file (202 ...) to create the *.ffe file?

It is very easy to calculate bond angles by RIETAN-94/orffe. Please
include line(s) using Function 2 of ORFFE just after line(s) using Function
201 in the input file of RIETAN-94. Then, you can get *.ffe containing
interatomic distances and bond angles in this order after running orffe.
Editing *.xyz to add lines containing Function 2 directly is an alternative
method.

When using the feature of nonlinear constraints on interatomic distances
and bond angles, please determine the initial value of a penalty parameter
T(0) in such a way that the penalty term (PF in the output) is 2-10% of the
weighted sum of squares (OF). If T(0) is too large, the least-squares
calculation may not converge.

RIETAN-94 can create a file for the PLOT program developed by Dr. Saktivel.
I have noticed that DMPLOT can read in this file if I modify only part of
one line. In the next version of RIETAN-94, combination of DMPLOT with
RIETAN-94 will be achieved. In addition, the unique input system, F3IS,
will be improved considerably.

********************************************
Fujio IZUMI
National Institute for Research in Inorganic Materials
1-1 Namiki, Tsukuba, Ibaraki 305, Japan
TEL: +81-298-51-3351 (ext. 378)
FAX: +81-298-52-7449
E-mail: izumi@nirim.go.jp
********************************************