Wear analysis of curved plate

Greetings team,

If the number of cycles is kept at 15, cycle increment 3 and wear smoothings step 5, the simulation is converging. But when i change to 15,000 cycles, 3000 cycle increment and 5 wear smoothing step, the model is giving me a non-positive jacobian error. I have tried to refine the mesh a number of time but to no avail. Kindly assist. Attached is the link to the file.

Looking only at the results the problem is at the elements on the symmetry plane.

In the third step, the boundary-displacement step, I would suggest adding a symmetric support to the symmetry plane on the bottom part. U3 = 0 mm.

oops! I have done as advised but unfortunately still getting the same error message after running the simulation.

I run the simulation, and the current settings are not ok.

After the wear step, the max computed wear depth is about 5 mm, but the plate thickness is less than that. So, the wear depth is too large to apply correctly to the mesh. So this will not work. The wear coefficient is way too low or the number of cycles is too high.