LGTM: Eliminate float-to-double overflow warning

LGTM complains if this calculation is done from inside the sqrt() call
because it sees the explicit cast to float and assumes that sqrt() is
intended to take a double. By adding an intermediate step it should be
clear to LGTM that the float version of sqrt is intended.

src/Mod/Mesh/App/Core/Grid.cpp

@@ -291,7 +291,8 @@ void MeshGrid::CalculateGridLength (unsigned long ulCtGrid, unsigned long ulMaxG
         else
             fAreaElem = fArea / float(_ulCtElements);
 
-        fGridLen = sqrt(fAreaElem * float(ulCtGrid));
+        float fRepresentativeArea = fAreaElem * static_cast<float>(ulCtGrid);
+        fGridLen = sqrt(fRepresentativeArea);
     }
 
     if (fGridLen > 0) {