uladkasach/finite-capacitance

Numerical solution of capacitance for a finite parallel plate capacitor of arbitrary plate specification.

finite-capacitance's Issues

Capacitance not being calculated correctly for ideal case, in center

Capacitance across a path between the two plates through the center of the plates should be equivalent to the solution in the ideal case, e_0 * A / d.

This is not so currently.

For A = 0.3^2 and d = 0.00005, expected C = 1.593*10^8, calculated C is different

Parallelization is not utilized to speed up computation

All of the calculations are very well translatable into map-reduce type calculations.

Calculating potential using electric field at accurate discretization takes too long

As seen in table : https://github.com/uladkasach/Finite-Capacitance/blob/master/z_dev_notes/issue_1/data_table.md

Although this method of calculating potential results in accurate results - these results are only accurate when the discretization is large enough. At that scale, (8k*8k = 64 Million elements per plate thickness per plate), calculating electric field for one position takes a very long time. Calculating the electric field 50 times to integrate over a path makes this method untenable.

It is possible to calculate potential directly right next to each plate and take the difference, since we are guaranteed in our model that the plates are finite. This would only require 2 iterations across all charge carrying elements (once for each position to calculate potential at.). This would be much faster and much more feasible.

Explicitly storing XY positions of conductor elements takes too much memory

Especially when considering that no arbitrary shapes have yet to be needed. This can be removed and replaced by calculating the positions of the conductor elements using the rectangular edge points.