[Bojan]

Hi everybody,

I have a "simple" problem which seems to be unsolvable.

I'm programming some mesh motion solver which is to be time dependent, but I have a problem when try to manipulate with a mesh point coordinate (here is the source of the problem):

movingPoints[curFace[pointI]] = meshPoints_[curFace[pointI]];

meshPoints_[curFace[pointI]] gives me (example line):

(-0.00649239 -0.010524 -0.0217109)

instead of

(-0.006492392086 -0.01052404258 -0.02171094868)

When I try to print the line using:

Info << "Point: " << Foam::setprecision(12) << meshPoints_[curFace[pointI]] << endl;

everything seems to be correct, but how am I going to use the full number length when performing further operations? Simply, I'm not able to proceed with mesh point coordinates which actually do not exist, because I get the Terminal message:

Aborted (core dumped)


Thanks in advance!

[Bojan]

It seems that the problem with number precision is caused by some additional preprocessor directive. Quite strange. Simple printing of the mesh points coordinates of turboFvMesh is working right without using Foam::setprecision(12), which here certainly points to the problem with some possibly unnecessary #include directive.