[tdof]

Hi all yall from good ol' Germany,
I'm currently writing my masters thesis with the subject of viscoelastic fluid flow of HDPE in an extrusion head. I'm no expert in CFD with only two years of experience using Acusolve, Abaqus CFD (lol) and OpenFOAM (Extend), but I think I'm no noob anymore. Unfortunately, I can't figure out how to further stabilize my current simulation so that it doesn't diverge after a short time. It's a real extrusion head geometry, so it's not possible to upload it here. However, you can access the initial/boundary conditions, numerical settings and the log file up to termination due to floating point exception.
https://1drv.ms/f/s!AqGKZzn3ghyumAOHjComcuvd9WiW
I used rather restricting relaxation factors of 0.2 for testing purposes, but to no avail. Also, I initialized the flow field by using pimpleFOAM and a newtonian, viscous fluid which stabilized the simulation a bit, but it starts to diverge after a certain time (~0.3s) nonetheless. I also halved the inflow velocity for testing purposes, but without any improvement. I don't know if this behaviour comes from the mediocre mesh (rather large tetras with hexa boundary layer created with Hypermesh) or the geometry itself, since there is a rather narrow outlet and complex middle part. The max Co number is around 10x the mean Co number, which is OK by my standards. It's a similar inner geometry to this, but a lot narrower: http://www.c-ramic.com.au/images/images_a/extrude.jpg Mind the spider in the middle part. Also, there is a 90° J-bend at the beginning of the geometry, and it's halved for symmetry purposes.
Previously, I studied generic extrusion dies with a regular hex mesh which all ran fine and I was able to examine various numerical settings, but with this geometry the optimum settings do not work. Does anyone have an idea how to get it running properly without setting the time step to 1e-6 or something similar? If I forgot anything and you need further information, just ask away.
If it's useful to anyone, heres the checkMesh output:

Mesh stats all points: 246780 live points: 246780 all faces: 1545724 live faces: 1545724 internal faces: 1424624 cells: 664924 boundary patches: 4 point zones: 0 face zones: 1 cell zones: 1
Overall number of cells of each type: hexahedra: 0 prisms: 310421 wedges: 0 pyramids: 231 tet wedges: 0 tetrahedra: 354272 polyhedra: 0
Checking topology... Boundary definition OK. Point usage OK. Upper triangular ordering OK. Face vertices OK. Number of regions: 1 (OK).
Checking patch topology for multiply connected surfaces ... Patch Faces Points Area [m2] Surface topology
fixedWalls 104605 53306 0.0512882 ok (non-closed singly connected)
sym 15911 11666 0.0070366 ok (non-closed singly connected)
inlet 324 230 0.000126968 ok (non-closed singly connected)
outlet 260 264 4.85937e-05 ok (non-closed singly connected)
Checking geometry... This is a 3-D mesh Overall domain bounding box (-3.282e-07 -0.459 -0.075) (0.028 0.009 0.028) Mesh (non-empty, non-wedge) directions (1 1 1) Mesh (non-empty) directions (1 1 1) Mesh (non-empty, non-wedge) dimensions 3 Boundary openness (1.09329e-15 9.40314e-17 3.6001e-16) Threshold = 1e-06 OK. Max cell openness = 3.93485e-16 OK. Max aspect ratio = 9.66468 OK. Minumum face area = 2.58494e-08. Maximum face area = 2.90099e-06. Face area magnitudes OK. Min volume = 9.36274e-12. Max volume = 1.64232e-09. Total volume = 0.000137297. Cell volumes OK. Mesh non-orthogonality Max: 64.3557 average: 18.0619 Threshold = 70 Non-orthogonality check OK. Face pyramids OK. Max skewness = 1.64549 OK.
Mesh OK.
End