mfisch
Hello
Assume I have several xdd's in an .inp file, as e.g:
xdd_1
str
a = param0 + param1 * temperature1 + param2 * temperature1^2;
xdd_2
str
a = param0 + param1 * temperature2 + param2 * temperature2^2;
xdd_3
str
a = param0 + param1 * temperature3 + param2 * temperature3^2;
How can I do that param1 and param2 are refined in xdd_1 and xdd_2 but only read in xdd_3?
- !param1 and !param2 in xdd_3 does also fix param1 and param2 in xdd_1 and xdd_2,
- =Get(param1) and =Get(param2) doesn't work,
- ...?
I guess this must me possible without rewriting a very long file...
alancoelho
Use
xdd_3
str
a = param0 + Constant(param1) * temperature3 + Constant(param2) * temperature3^2;
mfisch
But this is only updated at the beginning of the refinement, not after each cycle?
alancoelho
I'm therefore not understanding what you mean. Can you elaborate please.
mfisch
Is "constant" updated after each iteration or only when I accept the results, i.e. when the refinement has converged and I accept the results?
I am looking for something that is updated after each iteration but does not contribute to the refinement of
param0, param1, param2 in xdd_1 and xdd_2 but only to temperature3 in xdd_3.
So,
xdd_1: refinement of param0, param1, param2, temperature1.
xdd_2: refinement of param0, param1, param2, temperature2.
xdd_3: param0, param1, param2 are only read and temperature3 is refined.
each change in param0, param1, param2 in xdd_1 and _2 should directly be propagated to xdd_3.
alancoelho
I think I now understand.
The answer however is no. A parameter that is refined is treated as a variable in all xdds. This meas that its derivatives from all xdds are included in the normal equations of Ax=b.