Invited Speaker: Ming Li (on behalf of Tianjue Zhang)
Abstract: Proton beam with an average power of 5MW-10MW have important applications in particle physics towards the intensity frontier, as well as in the advanced energy, and material science. In 2019, a 2GeV/6MW CW FFA was proposed in China Institute of Atomic Energy (CIAE), which adopts a 100 MeV and 800 MeV Ring cyclotron for injection. From the perspective of the accelerator physics, the 2GeV FFA solution with 10 cells meets most of the requirements, except that the Vr=3 resonance are crossed. The technical requirements for the RF system are far from critical which is feasible in engineering. While the tolerance of the first harmonic, less than 1 Gs due to the resonance crossing, is too harsh for the magnet engineering. In 2020, the progress for the optimization design and construction of the 2GeV FFAG includes: 1) Physical design optimization. Making compromise between various physical design parameters and engineering parameters, solutions with different cell numbers and overall size are compared to determine the optimized physical design scheme, which is expected to be feasible in the engineering of the major systems, such as the magnet and RF system. 2) Pre-research of the high temperature superconducting (HTS) magnets. The physical, engineering design of the test-magnet and the HTS coils winding test have been completed. Through the confirmed technical route, the fabrication of the HTS coils has started and is planned to be finished in January 2021. In addition, the engineering design of the radial-varying-gradient formal magnet has also been completed and fabrication will begin soon. 3) RF system design and manufacturing. After the physical and engineering design of the RF system has been completed, we start the machining of the scale-down high frequency cavity with high Q-Value and large radial width, which is expected to be completed in February 2021. This report will outlined the status of the FFA design and the detail of the recent progress.
9 replies on “China Institute of Atomic Energy 2GeV, 6MW isochronous FFA proposal and R&D status”
Questions:
1) How does the FFA differ from a ring cyclotron?
2) 6MW machine must consider beam loss. Remember ,<1W/m losses for hands on maintenance. What features built into design to keep loss below 10^(-4) or smaller?
3) Can an HTS magnet quench? If it does quench, what happens?
4) Do you consider peak E-filed, peak B-field and multipactor or other arc discharge phenomena in the RF cavity?
5) Do you consider beam coupling to cavity HOMs?
6) Finally, the cavity in slides #35 and onwards is racetrack type. Why did you choose racetrack over boat type?
The video appears to be only 9:28 in length. It’s clearly not finished at the end. But I could read the slides.
It looks like an exciting project. (But I have same comment as Shane. Calling this FFAG rather than cyclotron is kind of arbitrary.)
There have been 2GeV cyclotrons proposed in the past. See “ASTOR” from PSI in the 1980s; this was a much more compact design. It seems that expanding the ring to such large size gives extra orbit spacing: Is that the reason it is so seemingly space-inefficient?
The cost of the +/-20 degree phase excursions is that the tolerances on magnet and frequency control are tighter. What do you pay if you try to tune the isochronism straight?
1) Actually the FDF lattice and spiral angle are adopted in this mahine, which could generate alternating focusing. So it better to call it FFA here.
2) We did not quantify the beam loss in this machine. But qualitatively, due to the arranged many cavities with high peak voltage, the turn number is only 50~60, and the turn separation at injection and extraction is very large, which could ensure that beam loss is very low. Much lower beam loss than the PSI ring cyclotron could be expected.
3) As shown in the slider 28, the working point of the superconducting coil is about 50% of the critical current Ic. The HTS magnet adopts large design margin to ensure the working reliability. And the quench protection circuit should be designed carefully. The quench problem should be further studied indeed.
4) The peak E fields are well below kilpatrick limit. Though peak H field may be low in absolute value, mechanical reinforcement to prevent Lorenz detuning are emphasized during design phase.
5) Several multipack regions of the resonanter have been identified using analytical method, numerical analysis will be carried out in the recent future to further address the issues.
6) Actually, in slides #35 the cavity is the boat type, not the racetrack type. We choose the boat type for this machine. The boat type has the highest qulity factor.
There are two reasons why we have made the machine so big. Firstly, long dirfts are adopted in this machine, so many cavities, injection and extraction elements could be arranged convinently. Secondly, as you said, large size gives enough extra orbit spacing for injection and extraction, which i think is very important for this machine. In a word, large turn separation and less turn numbers are very important for a high power cyclotron to reduce the beam loss.
For the tuning of the isochronism, each magnet could be machined to adjust the average field. There may be some difficulties in engineering. We have experiences on the mapping and shimming of the 400 ton magnet of our CYCIAE-100 machine. Besides, i think +/-30 degree phase excursions could be accepted. Intuitively, it is not an unsolvable problem.
I’ve seen some of the early GeV energy cyclotron designs. In their machines, it seems that the crossing of integer resonance is not a big problem. These designs seemed to underestimate the destructive effect of integer resonance crossing. While in our 2 GeV machine, very small 3rd harmonic field could generate a blow up of the radial beam envelop. So we did a lot of work to avoid the resonance crossing of Vr = 3 resonance.
I have a question relative to Shane’s question 1) (“How does the FFA differ from a ring cyclotron?”) (just for my education):
Is there any such cyclotron w/ alternate gradient dipole triplet cell, or even doublet FD cell, in operation?
Followup question. Did you investigate the nu_r=5/2 resonance? Fifth harmonic can also arise from uneven cavity distribution. Magnetically, it is also intrinsic as 10/4. But this effect is probably negligible as it was for TRIUMF’s 6/4.
Thank you. There is not such cyclotron in operation. I think it is better to call it isochronous FFA or CW FFA.
Thank you for your comments.
Uneven cavity distribution did not impact the beam envelop through our simulation. But gradient of fifth harmonic magnetic field ( like 2 Gs/cm B5 in a radial range of 30cm) indeed give a radial beam envelop growth(about 50%). But i think it is not so easy to generate such a large field gradient in a wide radial range, right? We must study this point clearly further, taking into account engineering considerations.
Resonance with order > 2 did not influence the beam from our simulation, including 4Vr=10.
Further on the nu_r=3 resonance: I estimate from the known formula and get amplitude growth three times larger than in your simulation! So I need more detail. How was the third harmonic error field applied in the calculation? What is the change in tune per turn when crossing? (Of course I mean the 2019 design, not the one that avoids it.)