Singularity of surface currents from bempp

[Sangwoo Kang]

Dear all 

Hello. 

I try to reconstruct the surface currents J=n x H, M = - n x E of target using bempp.

But I found that the result has a singularity in J even though the scattered fields have the accuracy results from J and M. 

                                         J                                                                                                    M

This is the result of the example "maxwell_dielectric.ipynb"  given in the bempp website. 

The result is shown by 

J for sol[1].plot(mode='vecrtices', transformation='abs_squared') 

M for sol[0].plot(mode='vecrtices', transformation='abs_squared') 

I used both bempp-cl and bempp3.3.4, but this phenomenon was still observed. 

Do you have any idea for the reason and method to handle it?

Thank you in advance.

BR,

Sangwoo Kang