Study of High-gradient Structures for the Linear Collider

对直线对撞机主直线加速器的加速结构进行了实验和理论研究。该结构为行波型，采用11.4 GHz X波段射频功率驱动。加速梯度希望高达100 MV/m。在本财政年度，我们制造了具有非常精确和精细表面光洁度的无氧铜的原型结构。然后在KEK使用一台30兆瓦的速调管进行了高功率测试，这是在KEK开发的。我们能够在500小时的条件反射后达到高梯度。在很高的射频电场下，我们还对表面发射产生的暗电流进行了详细的测量。为了获得高梯度，高功率速调管是绝对必要的。然而，随着工作频率的提高，速调管的制造技术变得极其困难。因此，在S频段(2.86 GHz)进行了高梯度实验，在该频率下，大功率速调管容易…更多可用的y。事实上，在这个频率下，我们成功地实现了80 mV/m以上的平均梯度，并测试了束流加速度。然而，对于TeV能量线性对撞机，射频频率最低应为X波段(11.4 GHz)，否则总交流功率将变得不切实际地高。KEK对X波段大功率速调管的研制由来已久，最近我们成功研制出了第一个可输出30 mW射频功率的管。这项实验是用这个管子进行的。欧洲核子研究中心对TeV直线对撞机有不同的想法。特别是，他们正在考虑由自由电子激光产生的K波段(30 GHz)频率。它的实现仍然是遥远的未来。另一方面，他们正在开发能够制造如此高梯度的加速结构的技术。因此，他们希望尽快测试这种结构，即使在较低的频率下也是如此。因此，在KEK开始了一项联合实验。在这个实验中，CERN和KEK制造了相同尺寸的结构。主要规格为：单元数20个，耦合单元2个，单元长度2.9 mm，单元相移2pi/3，恒阻抗结构，光束口径6 mm，单元直径10.4 mm，盘厚1 mm。Q值为6500，衰减为1.54/m，阻抗为104MOMEGA/m，光束口径上的最大表面场是平均梯度的3.9倍。制作过程中最关键的一点是在光束光圈上获得足够的光滑度。这是通过精心选择钻石钻头来实现的。我们还注意避免金属和有机粉尘落在结构内部。它们通常是通过引起强放电来限制可达到的最大斜率的严重原因。测试是在速调管重复频率50赫兹下进行的。RF脉冲长度通常为100 ns。KEK结构在6000万脉冲后达到70 mV/m，而CERN结构在900万脉冲后达到100 mV/m。在欧洲核子研究中心的案例中，我们还使用了雪橇系统来提高射频功率。观察到的暗电流使用改进的Fowler-Nordheim公式进行分析。其中重要的量度是场增强因子β。对于这两种结构，我们得到了大约50的β。对于KEK结构，暗电流的绝对值被发现相当高，需要进一步的研究来解决这个问题。较少

Experimental and theoretical studies have been carried out for accelerating structures of the main linac of linear colliders. The structure is of a traveling wave type driven with 11.4GHz X-band RF power. The accelerating gradient is desired to be as high as 100MV/m. In this fiscal year, we fabricated a proto-type structure of oxygen free copper with a very accurate and fine surface finish. It was then high power tested at KEK by using a 30MW klystron which had been developed at KEK. We were able to reach the high gradient after a 500 hour conditioning. We also made detailed measurements of dark currents due to surface emission under the very high RF electric field.In order to achieve a high gradient, a high power klystron is absolutely necessary. However, the fabrication technology of klystrons becomes extremely difficult as the operation frequency raises. Therefore high gradient experiments were carried out at the S-band(2.86GHz), at which frequency the high power klystrons are easil … More y available. In fact at this frequency we achieved over 80MV/m average gradients successfully and also tested beam acceleration too. For the TeV energy linear collider, however, the RF frequency should be an X-band(11.4GHz) one at the lowest, since otherwise the total ac power becomes impractically high. KEK has been developing X-band high power klystron for a long time, and recently we succeeded to get a first tube which can deliver 30MW RF power. This experiment is carried out by using this tube.CERN has a different idea for the TeV linear collider. Particularly they are considering a K-band(30GHz) frequency generated by a FEL. Its realization is still far in the future. They are, on the other hand, developing technologies to fabricate accelerating structure which is capable of such high gradients. They are therefore wanting to test the structure as soon as possible even at lower frequencies. Hence a joint experiment started at KEK.In this experiment CERN and KEK made structures of the same dimensions. The main specification is : number of cell 20, coupler cell 2, cell length 2.9mm, phase shift per cell 2pi/3, constant impedance structure, beam aperture 6mm, cell diameter 10.4mm, disk thickness 1mm. The Q value is 6500, attenuation is 1.54/m, impedance is 104MOMEGA/m. The maximum surface field over the beam aperture is 3.9 times the average gradient. The most critical point in the fabrication was to achieve a sufficient smoothness on the beam aperture. It is accomplished by carefully choosing diamond bits. We also paid attention to avoid metallic and organic dusts falling inside the structure. They are usually a serious cause to limit the maximum attainable gradient by inducing heavy discharges.The test was carried out at the klystron repetition rate 50Hz. The RF pulse length was typically 100ns. The KEK structure reached 70MV/m after 60million pulses, While the CERN one did 100MV/m after 9 million pulses. In the CERN case, we also used a SLED system to boost the RF power.Observed dark currents were analized by use of a modified Fowler-Nordheim formula. The important measure therein is the field enhancement factor beta. For ther both structures, we obtained similar values around 50 for beta.The absolute value of the dark current was found to be rather high for the KEK structure and further studies are necessitated to resolve this. Less