"Non positive Jacobian Determinant" ????? in a Centrif type analysis

[Westcliffe01]

I have been doing pretty good so far with my flywheel development, so I tried to analyse the flex plate that has been cracking so far. No problem with the model. But I cannot get the FEA to complete. It reports "nonpositive jacobian determinant" in several elements ? I have tried re-meshing with finer mesh ? Any idea what is up with this error ? Thanks.
The file is too big to attach, 30.xxx Mb otherwise I would. Nothing secret.

0.0: Check dependencies...
5.9: Write completed.
0.0: CalculiX binary: C:/Program Files/FreeCAD 0.20/bin/ccx.exe
0.0: CalculiX input file: C:/Users/Keith/Pictures/Arion Lightning WV Jab 3300/Viking 110 708HU/Flex plate\SolverCcxTools\FEMMeshNetgen.inp
0.0: Run CalculiX...
0.0: Starting CalculiX...
0.0: CalculiX is running...
22.6: CalculiX stopped.
22.6: ************************************************************
CalculiX Version 2.17, Copyright(C) 1998-2020 Guido Dhondt
CalculiX comes with ABSOLUTELY NO WARRANTY. This is free
software, and you are welcome to redistribute it under
certain conditions, see gpl.htm
************************************************************
You are using an executable made on Thu Jul 23 21:43:50 CEST 2020
The numbers below are estimated upper bounds
number of:
nodes: 462313
elements: 390008
one-dimensional elements: 0
two-dimensional elements: 0
integration points per element: 4
degrees of freedom per node: 3
layers per element: 1
distributed facial loads: 0
distributed volumetric loads: 2
concentrated loads: 0
single point constraints: 636
multiple point constraints: 1
terms in all multiple point constraints: 1
tie constraints: 0
dependent nodes tied by cyclic constraints: 0
dependent nodes in pre-tension constraints: 0
sets: 6
terms in all sets: 1969846
materials: 1
constants per material and temperature: 2
temperature points per material: 1
plastic data points per material: 0
orientations: 0
amplitudes: 3
data points in all amplitudes: 3
print requests: 1
transformations: 0
property cards: 0
STEP 1
Static analysis was selected
Decascading the MPC's
Determining the structure of the matrix:
number of equations
1386303
number of nonzero lower triangular matrix elements
50148483
Using up to 8 cpu(s) for the stress calculation.
Using up to 8 cpu(s) for the symmetric stiffness/mass contributions.
*ERROR in e_c3d: nonpositive jacobian
determinant in element 183838
*ERROR in e_c3d: nonpositive jacobian
determinant in element 183838
*ERROR in e_c3d: nonpositive jacobian
determinant in element 183838
*ERROR in e_c3d: nonpositive jacobian
determinant in element 183838
*ERROR in e_c3d: nonpositive jacobian
determinant in element 285497
*ERROR in e_c3d: nonpositive jacobian
determinant in element 285497
*ERROR in e_c3d: nonpositive jacobian
determinant in element 285497
*ERROR in e_c3d: nonpositive jacobian
determinant in element 285497
22.6: CalculiX execute error:
22.6: Loading result sets...

Flex plate crash.JPG (298.4 KiB) Viewed 1535 times

[NewJoker]

Here you can find some tips regarding this error: https://wiki.freecadweb.org/FEM_MeshGmshFromShape#Notes

You could share the model using some hosting website and just paste the link here. Also, if you clear the mesh the file should be much smaller.

Did you try with a coarser mesh and Gmsh mesher ?

[Westcliffe01]

Simulation (after deleting all the lightening holes, thus stronger)

flex-plate-04.JPG (75.83 KiB) Viewed 1370 times

Cracks on actual part at 350 hours.

Drive plate cracks-1.jpg (146.58 KiB) Viewed 1370 times

[PMMA]

Hey,

The cracks you detected are very clear.
Is there a weld near the holes?

Be aware that the weld brings a metallurgic notch effect into your construction, which your FEA wont calculate in.

Maybe you should take a closer look to your WPQR and the hardness-testing.
Is it austenitic or why you doing PT?

Anyway your calculation could be much smaller by using symmetry.

[Westcliffe01]

There is no weld, in fact the part is stamped from a single sheet. The automotive part is zink plated, the modified part for aviation has paint applied and the central area where it attaches to the crankshaft is masked to keep the steel bare. The edge of the masked area there is a paint ridge, it is not a defect in the metal. The crack that formed is from a gyroscopic bending moment that essentially behaves like a rotating bending moment. The material of the flex plate is usually attached to the torque converter so that the engine torque can be transferred. The starter ring gear is usually also carried by the torque converter. In the aviation application, the ring gear was welded to the flex plate which is not supported by the torque converter. Obviously the material thickness of the flex plate is not fit to carry the gyroscopic loads of the ring gear. So far we have seen 3 examples fail in less that 350 hours.