In the TRA simulation of an ideal lossy line, the loss seems to be miscalculated. If one puts A = 1dB/m and makes a 1-m-long transmission line, the amplitude of a rectangular pulse at the end of line is reduced only by 0.5 dB with respect to incident voltage pulse at the beginning of the line, instead of 1 dB. The same goes for losses of A = 20 dB/m, where the pulse demonstrates only 10 dB loss on 1m line instead of 20 dB.
The AC analysis works fine though. Losses of sine signals in exactly the same transmission line follow the theory.