[gfilip]

Hello,

I am trying to continue a parallel simulation started in 1.7.x (before split cyclics) with 2.1.x. I ran the foamUpgradeCyclics utility in parallel with no errors, but once I try to launch the simulation, I get the following error and warnings:

Create time

Create mesh for time = 165.5

--> FOAM Warning :
    From function cyclicPolyPatch::calcTransforms()
    in file meshes/polyMesh/polyPatches/constraint/cyclic/cyclicPolyPatch.C at line 348
    Specified separation vector (0 0 1) differs by that of neighbouring patch (0 0 1) by more than tolerance 1.7944e-05
patch:sides_half0 neighbour:sides_half1
--> FOAM Warning :
    From function cyclicPolyPatch::calcTransforms()
    in file meshes/polyMesh/polyPatches/constraint/cyclic/cyclicPolyPatch.C at line 348
    Specified separation vector (0 0 1) differs by that of neighbouring patch (0 0 1) by more than tolerance 1.7944e-05
patch:sides_half1 neighbour:sides_half0
--> FOAM Warning :
    From function cyclicPolyPatch::calcTransforms()
    in file meshes/polyMesh/polyPatches/constraint/cyclic/cyclicPolyPatch.C at line 368
    Specified separationVector (0 0 1) differs from computed separation vector 1((-7.10543e-15 2.22045e-16 -1))
This probably means your geometry is not consistent with the specified separation and might lead to problems.
Continuing with specified separation vector (0 0 1)
patch:sides_half1 neighbour:sides_half0
Reading field p

Reading field U

Reading/calculating face flux field phi

and in the .o file:

[62]
[62]
[62] --> FOAM FATAL IO ERROR:
[62] essential value entry not provided
[62]
[62] file: /nobackup/gfilip/cylinderLES/Re1E06_LES/processor62/165.5/phi::boundaryField::sides_half0 from line 320181 to line 320181.
[62]
[62]     From function fvsPatchField<Type>::fvsPatchField
(
    const fvPatch& p,
    const DimensionedField<Type, surfaceMesh>& iF,
    const dictionary& dict
)
[62]
[62]     in file lnInclude/fvsPatchField.C at line 110.
[62]
FOAM parallel run exiting
[62]
[56]
[56]
[56] --> FOAM FATAL IO ERROR:
[56] --------------------------------------------------------------------------

and this repeats for all other processor dirs. I checked the phi files and they were converted to split cyclics. Any clues?