First Steps Toward the Development of a Straight-Line Reference Alignment System for Future Accelerators at CERN Using Pseudo-Nondiffracting Layer Beams

TL;DR

This work addresses the need for robust, long-distance straight-line alignment in accelerator facilities by developing Layer Beams (LBs), a class of pseudo-nondiffracting beams able to generate parallel reference planes for multi-point measurements. The authors implement a hollow sensor system and fiducialisation to quantify plane quality and relative displacements, validating performance over a 2 m path against laser-tracker references and identifying key noise sources such as CMOS cover glass reflections and dark noise. They also demonstrate a radiation-hard sensor concept using an optical fiber matrix, maintaining sub-micrometer alignment accuracy (RMSE < 1.3 μm) under radiation exposure. Collectively, the results support LB-based alignment as a viable, non-obstructive alternative to traditional shutter-based optical references, with clear paths for long-range deployment (140 m benches) and integration with existing accelerator subsystems.

Abstract

This paper presents experimental results that allow for the performance evaluation of a straight-line reference alignment system based on pseudo-nondiffracting Layer beams. Sensors, developed specifically for this system, feature four linear CMOS chips and a square aperture. This allows for simultaneous measurements along the beam path without disrupting the laser reference. Measurements, conducted over a distance of 2 m from the first to the last sensor, were compared with a laser tracker measurement to assess the sensor performance. The alignment reference generated by the Layer Beams exhibited a repeatability and reproducibility root-mean-square error (RMSE) of less than 30 $μ$m. The relative alignment precision for a known displacement was validated with a standard deviation of 4.3 $μ$m. The results highlight the underlying sources of noise, which are induced mainly by the cover glass, the protective film of the pixels, and the dark noise of the CMOS chips. Solutions to address these challenges are proposed. Additionally, a proof-of-concept for future development of a radiation-hard sensor utilizing optical fiber matrices is demonstrated. The RMSE of the reference position detection introduced by the fiber matrix remained below 1.3 $μ$m. This would allow the sensor to be used reliably in high-radiation environments typical for accelerator facilities. This study serves as a foundational step toward developing a robust straight-line reference alignment system based on pseudo-nondiffracting Layer beams intended for deployment in the accelerator facilities.

Figures (10)

• Figure 1: LB generator consisting of a dimension, phase, and projection lens with an illustration of a generating wavefront that dictates the caustic properties of the beam.
• Figure 2: Horizontal LB imaged on a camera at the position of the first sensor - 1.7 m away from the generator. Diffraction patterns caused by dust on the CMOS chip were removed in post-processing.
• Figure 3: Illustration of an alignment reference plane created by a LB.
• Figure 4: Illustration of an alignment system based on LBs.
• Figure 5: CAD model of the hollow sensor developed for the detection of LBs. A photo of the sensor can be found in supplementary materials.