|Status:||Backlog||Start date:||17 Jul 2013|
Instead, the following happens:
- kz=0 in the top layer of a multilayer sample with more than one layer: this means zero glancing angle and we take T0=1 and R0=-1 (as this represents the correct limit). All boundary conditions at the interfaces are put to zero;
- kz=0 in the only layer: in this case, we only have Born approximation and T=1 and R=0;
- kz=0 in a layer, other than the top layer of a multilayer sample: while the correct boundary conditions will be calculated, the used profile is a constant one and T=boundary value of wavefunction and R=0.
#4 Updated by herck about 6 years ago
The real problem does not reside in calculating the layer coefficients (amplitudes of R and T) but in the fact that included particles in the layer no longer scatter according to the Fourier transform of the shape function (there is no longer an exponential wave inside the layer but a linearly decreasing one).
#12 Updated by wuttke about 5 years ago
- Status changed from Sprint to Backlog
- Priority changed from High to Normal
- Target version deleted (
- Parent task changed from #983 to #1102
For small glancing angles, it is necessary to carefully consider the effects of finite beam width, finite sample extension, and multiple scattering (#1102). Only if this is properly resolved it will make sense to come back to the very special, practically unimportant case f=0.