A unique platform for a Tera Hertz Transmission Line design for a superradiant FEL is present. The smart line is controlled by Artificial Intelligence (AI) intended for a wide tunable broad-spectrum THz radiation propagation. The main goal is to transfer radiation in the most efficient way. A 3D analysis and diagnostic of radiation space-frequency tool was developed. The AI changes the functions of the mirrors in such a way that all the reflected rays will reach the target. The rays represent the electromagnetic field similar to a light field. The representation of the field in terms of rays was carried out using the Wigner Distribution Function. It allows describing the dynamics of field evolution in future propagation. This in turn helps with the initial design of the transmission line and facilitates the use of a Ray Tracing method for future processing. Thus, working in the linear and non-linear regimes. The Ray Tracing method and code is greatly enhanced using parallel processing with graphics cards.

This project describes different techniques to manufacture THz mirrors with arbitrary surfaces. The research is part of the development of THz transmission line for the compact FEL-THz accelerator. As an initial phase flat mirrors were 3D printed with FFF (Fused Filament Fabrication) and SLA (Stereolithography Apparatus). The impact of material, layer height and layer direction to mirror’s surface quality was exanimated. In addition, various metal coating was tested, for example vacuum evaporation and metal foil. The 3D printed flat mirror’s reflection was measured in TDS (Time Domain Spectroscopy) at 1–5 THz and compared with aluminum metal plate and glass silver coated mirror. The results approve sufficient surface and coating quality. Further research is manufacture off-axis parabolic mirrors, validate with a beam profiling and manufacture arbitrary surface mirrors optimized to the current accelerator by machine learning.