The timing couldn’t have been better for Bozo Richter, PhD student at the Technical University of Hamburg and at DESY. For his PhD thesis in control engineering, he was supposed to develop a system to make the planned upgrade of the European XFEL accelerator more robust against external mechanical disturbances. This upgrade is proposed for the next decade, but nevertheless the system needed to be tested. It so happens that there is a particle accelerator at the SLAC National Accelerator Laboratory in California, USA, which uses the same technology as the European XFEL. Thanks to years of collaboration between SLAC, DESY and other research centres, Richter was able to test his system so successfully on the US accelerator that it is now also being used there.

But let’s begin at the beginning. Particle accelerators are usually rather large, impressive machines packed with high-tech components, but as is so often the case, the devil is in the detail. There are various types of particle accelerators, the design and technology of which depend on what they are supposed to accelerate and how the research community intends to use them. At the European XFEL and its ‘sister’ facility at SLAC – with the similarly catchy name LCLS II – it’s electrons, which are used to generate particularly high-quality X-rays; these, in turn, can be used to examine all manner of things – from new materials to chemical processes in real time.

This type of acceleration often relies on a technology in whose development DESY played a central role and, as PhD student Bozo Richter demonstrates, continues to do so: superconducting radio-frequency (SRF) technology. The key components are the cavities, one-metre-long Baumkuchen lookalikes, just with a silver glaze instead of chocolate. They are made of the element niobium, which has the neat property of conducting electricity without resistance once temperatures are close to absolute zero. This material is formed into several perfectly identical cells, into which high-frequency radio waves are fed. Due to the cavitiy's special shape, these waves are reflected back and forth in such a way that an extremely strong electric field is generated down the middle. If electrons are then fired into the cavity, they are accelerated to extreme energies by this electric field – ideally with no losses at all, thanks to superconductivity.

However, it is not only superconductivity that influences the cavity’s efficiency. The radio wave inside it is tuned with extreme precision, much like a perfectly tuned musical instrument. The cavities must not deviate by even a micrometre from their intended shape, nor contain any impurities – anyone who has ever heard an out-of-tune piano will realise that this would be just as detrimental to acceleration as an out-of-tune piano is to our ears. But what if something from the outside affects the ideal shape? Even the slightest vibration or shock could trigger a minor deformation, which would then disrupt the wave inside and compromise the electric field. And a particle accelerator of this kind has a great many components that can vibrate, hum, make noises or rattle. If the high-frequency system has to constantly compensate for these deviations, the required amplifier power increases – the cavity becomes less efficient and needs significantly more energy.

This is precisely where the system developed by Bozo Richter and his PhD colleagues Radosław Rybaniec and Andrea Bellandi comes into play. It detects external disturbances and readjusts the cavity through micro-changes in shape so that the field inside remains unaffected. It is called “NANC”, short for “Narrowband Active Noise Control”, and is an algorithm which, in addition to existing noise suppression systems, monitors the situation, continuously analyses the disturbances and adjusts the counter-signal. “For example, it detects when a pump suddenly has to work a little harder and then returns to its normal load,” explains the PhD student. “The algorithm can adapt the cavity’s compensation directly to the changes in the pump.”

The system was tested on a test bench at DESY and during live operation at a test station on the LCLS-II accelerator. “It was incredibly exciting to be able to observe in real time the impact the algorithm had on accelerator operation,” says Richter, who spent some time at SLAC for this purpose and worked with the team on site. “A real win-win situation: I was able to test part of my PhD thesis and, as a result, LCLS II ran more smoothly! There are a few areas where particularly high levels of vibrations were identified only after commissioning of the accelerator. One of the cavities in that area would not have been operational without NANC, because the existing amplification power would not have been sufficient. With NANC, it worked.” In some cases, several NANC components were able to significantly reduce detuning and improve acceleration efficiency by a factor of two.

The algorithm is particularly important for future accelerators (or their upgrades) that are intended to run with very high duty cycles or continuous acceleration, and for which rapid regulation is crucial to their overall performance. By then, Richter will probably not only have his PhD under his belt, but also one or two postdoc contracts. He is already working on ideas for how he can further improve his system and extend it to a wider range of frequencies.

Incidentally, the Hamburg native is a prime example of the wide range of opportunities and roles offered by research centres such as DESY. “I’m from Hamburg and have always found DESY absolutely fascinating because of the physics and research involving photons. But there’s also so much engineering involved in the facilities, where mechanical engineers and mechatronics engineers like me can make important contributions,” says Richter.

This work was funded as part of the European XFEL’s research and development programme. The project also received funding from the European Union’s ‘Horizon Europe’ research and innovation programme under grant agreement No. 101131435.

 

Further information:

B. Richter, J. Branlard, A. Bellandi et al., Narrowband active noise control for continuous wave operation of TESLA cavities, Proceedings of the 17th International Particle Accelerator Conference 2026, https://www.ipac26.org/prepress/pdf/WEP6149.pdf

With support from iSAS - Innovate for Sustainable Accelerating Systems