69mancl
Subscriber
Hello,nthanks for answers. The setup is default so 30% refinment and 10 passes. Aimed error is 1%. nI've performed simulations with more passes and with sharp edges it can never converge as you can see in following table:nIf i fillet the edges then the situation has similar problems, but due to round edges and small details between the dielectric and the electrodes, the mesh also has to be more detailed and my computer is unable to do more than 11 passes. But as you can see in the next table, number of passes is not the problem.nWhen i remove the air dielectric, so only the air from area makes the dielectric, then i converge quite fast. I get the same result when i make dielectric less thick. In the next table you can see convergence without the dielectric model. nnAnd in the next picture you can see the convergence for less thick dielectric and how the model looks in detail. nSo the problem is in the small sharp edges. But im not sure why. nThe setting was the same for all the simulation, only the number of passes varies. n