Resolves bug-report http://bugs.openfoam.org/view.php?id=2273

fvPatchFields: Constructors from dictionary now call the corresponding constructor of the fvPatchField base-class
to ensure 'patchType' is set as specified.

Required substantial change to the organization of the reading of the
'value' entry requiring careful testing and there may be some residual
issues remaining.  Please report any problems with the reading and
initialization of patch fields.

Resolves bug-report http://bugs.openfoam.org/view.php?id=2266

e.g. for the cavity tutorial the moving wall patch can be specified in
terms of the block vertices as before:

boundary
(
    movingWall
    {
        type wall;
        faces
        (
            (3 7 6 2)
        );
    }
    .
    .
    .

or the new specification of the face as block 0, block face 3:

boundary
(
    movingWall
    {
        type wall;
        faces
        (
            (0 3)
        );
    }

Resolves bug-report http://bugs.openfoam.org/view.php?id=2264

Patch contributed by Mattijs Janssens
Resolves bug-report http://bugs.openfoam.org/view.php?id=1815

Resolves bug-report http://bugs.openfoam.org/view.php?id=2239

linearUpwind: Evaluate the gradient of each component of the field to provide support all field types

Also reduces peak-storage as it now generates a volVectorField for each component
rather than the gradient of the field type.

derivative of the stress tensor.  See
http://en.wikipedia.org/wiki/Upper-convected_Maxwell_model

The model includes an additional viscosity (nu) from the transport
model from which it is instantiated, which makes it equivalent to the
Oldroyd-B model for the case of an incompressible transport model.
See https://en.wikipedia.org/wiki/Oldroyd-B_model

Renamed the original 'laminar' model to 'Stokes' to indicate it is a
linear stress model supporting both Newtonian and non-Newtonian
viscosity.

This general framework will support linear, non-linear, visco-elastic
etc. laminar transport models.

For backward compatibility the 'Stokes' laminar stress model can be
selected either the original 'laminar' 'simulationType'
specification in turbulenceProperties:

    simulationType laminar;

or using the new more general 'laminarModel' specification:

    simulationType laminar;

    laminar
    {
        laminarModel        Stokes;
    }

which allows other laminar stress models to be selected.

Individual inward-pointing faces are checked and if all faces are
inward-pointing the block is inside-out.  These errors are fatal and the
message indicates which block the error occurs in and where in the
blockMeshDict the block is defined.

Resolves bug-report http://bugs.openfoam.org/view.php?id=2248

Patch provided by Mattijs Janssens
Resolves bug-report http://bugs.openfoam.org/view.php?id=2226

Now works correctly for an arbitrary number of bodies
Resolves bug-report http://bugs.openfoam.org/view.php?id=2211

rigidBodyMeshMotionSolver: experimental nDoF mesh-motion solver supporting the displacement-based elliptic solvers

Specification for the tutorials/multiphase/interDyMFoam/ras/floatingObject case:

dynamicFvMesh       dynamicMotionSolverFvMesh;

motionSolverLibs   ("librigidBodyMeshMotion.so" "libfvMotionSolvers.so");

solver             rigidBodyMotionSolver;

rigidBodyMotionSolverCoeffs
{
    report          on;

    meshSolver
    {
        solver displacementLaplacian;

        displacementLaplacianCoeffs
        {
            diffusivity inverseDistance (floatingObject);
        }
    }
.
.
.

Resolves bug-report http://bugs.openfoam.org/view.php?id=2244

Renamed 'omega' to 'FSDomega' to avoid a clash with the k-omega
turbulence models.

Resolves bug-report http://bugs.openfoam.org/view.php?id=2237

Patch provided by Mattijs Janssens
Resolves bug-report http://bugs.openfoam.org/view.php?id=2240

Patch contributed by Mattijs Janssens

Now the postProcess utility '-region' option works correctly, e.g. for
the chtMultiRegionSimpleFoam/heatExchanger case

postProcess -region air -func "mag(U)"

calculates 'mag(U)' for all the time steps in region 'air'.

Patch provided by Mattijs Janssens
Resolves bug-report http://bugs.openfoam.org/view.php?id=2226

algebraicPairGAMGAgglomeration: agglomerate based on the maximum of the upper and lower coefficients
Patch provided by Mattijs Janssens
Resolves bug-report http://bugs.openfoam.org/view.php?id=2234

using a run-time selectable preconditioner

References:
    Van der Vorst, H. A. (1992).
    Bi-CGSTAB: A fast and smoothly converging variant of Bi-CG
    for the solution of nonsymmetric linear systems.
    SIAM Journal on scientific and Statistical Computing, 13(2), 631-644.

    Barrett, R., Berry, M. W., Chan, T. F., Demmel, J., Donato, J.,
    Dongarra, J., Eijkhout, V., Pozo, R., Romine, C. & Van der Vorst, H.
    (1994).
    Templates for the solution of linear systems:
    building blocks for iterative methods
    (Vol. 43). Siam.

See also: https://en.wikipedia.org/wiki/Biconjugate_gradient_stabilized_method

Tests have shown that PBiCGStab with the DILU preconditioner is more
robust, reliable and shows faster convergence (~2x) than PBiCG with
DILU, in particular in parallel where PBiCG occasionally diverges.

This remarkable improvement over PBiCG prompted the update of all
tutorial cases currently using PBiCG to use PBiCGStab instead.  If any
issues arise with this update please report on Mantis: http://bugs.openfoam.org

Patch contributed by Mattijs Janssens
Resolves bug-report http://bugs.openfoam.org/view.php?id=2235

This allows freedom in the naming of patches when patchGroup-based
boundary specification is not used.

Patch contributed by Mattijs Janssens

Patch contributed by Mattijs Janssens
Resolves bug-report http://bugs.openfoam.org/view.php?id=2227

Resolves bug-report http://bugs.openfoam.org/view.php?id=2225

This reverts commit 12ee0179.

If this clash is allowed boundary conditions set by wildcards are
processed incorrectly.

Patch contributed by Bruno Santos
Resolves bug-report http://bugs.openfoam.org/view.php?id=2218