电子信息工程外文翻译参考文献
- 格式：doc
- 大小：525.00 KB
- 文档页数：14
电子信息工程外文翻译参考文献(文档含中英文对照即英文原文和中文翻译)译文：利用修改后的迈克耳孙干涉仪进行长度测量的初步结果摘要:基于飞秒加速器的装置，该装置建造在上海应用物理研究所(SINAP),最近一个经修改后的远红外迈克耳孙干涉仪通过光学自相关方法，已经被用来测量电子光束的长度。
相比较于之前常规的迈克耳孙干涉仪,我们使用一个空心回射器而不是一个平面反射镜的反射镜。
本文将为大家介绍实验设置和长度测量的结果。
关键词：飞秒线性加速器，长度串，干涉仪，空心回射器1 介绍最近关于电子脉冲压缩的实验产生高峰值电流和亮度飞秒电子串。
关于短束源自于高质量光束的潜在应用要求这方面一起了广泛兴趣。
高质量的核物理加速器,自由电子激光器驱动加速器,下一代线性对撞机,第四代光源都需要短时间光束脉冲。
同时,在进程中对诊断的短电子串的研究也起了重要作用。
有几种已经使用或正在开发的方法去测量短电子串的长度。
这些一般分为两类:频域方法和时域方法。
众所周知，在时域测量长度的方法中使用条纹相机，条纹相机已经证实是限于串长度超过200 fs ，此外,条纹相机昂贵并且测量系统复杂。
相对于时域测量方法，频域测量使用相干过渡辐射(CTR )从金属箔在测量飞秒脉冲的短电子中已经显现出前景。
本文我们首先从短电子串方面给出了基于一代的高强度相干渡越辐射的理论和试验研究,然后讨论该方法基于相干渡越辐射测量束飞秒的长度,并从改进电子实验装置给出了串长度测量的结果。
最后,我们分析了空气湿度对串长度测量的影响，并且阐释了对未来研究的计划。
2 理论背景2.1 相干渡越辐射源自于相对论性电子串辐射如同步加速器辐射跃迁辐射等，本质上有较广的范围，如果辐射的波长短于电子串长度，这个阶段的辐射电子不同于彼此，所以辐射是不连贯的。
另一方面,如果波长较长的串长度，辐射是连贯的并且辐射强度的平方成正比每串数字电子。
光谱强度发出一束N 粒子：11()()(1)()|()|tot I NI N N I f λλλλ=+- （2-1）这里1()I λ是靠单电子辐射的强度，()f λ是串形成因素，这是傅里叶变换的规范化的电子密度分布()S Z 。
对于一个相对横向尺寸小于长度的物体，形状因子就为()()exp[2()]f S Z i n z dz λπλ=⎰（2-2）其中n 是单位向量从群到观测点的中心，z 是电子相对于堆中心的位置向量。
显然,测量的辐射谱将通过傅里叶变换给出形成因素和电子密度分布。
2.2 电子串长度测量电子串长度往往是通过自相关的CTR 信号与一个迈克耳孙干涉仪进行测量，干涉仪是由一分束器，一个固定的平面反射镜和一个可移动的平面镜组成。
光进入迈克耳孙干涉仪是被光分束器分成两部分，这两部分在两个不同的方向运行，并被镜子反射回来。
经过反射后的两个辐射脉冲组合成一个戈利检测器来测量光强度。
得到的干涉图是通过测量探测器信号在两个分路的差异作为路径功能。
测量能量的调配辐射脉冲来自于固定镜子，()fix E TRE t =，从可移动的镜子辐射延迟的时间/c δ，(/)move E RTE t c δ=+。
这里()R R ω=和()T T ω=是反射和透射系数的分束器。
强度测量探测器可以表示为（2-3）其中δ是光速差，c 是光速。
或者,可以在频域从动臂通过添加一个额外的相位差/i c eωδ-和角频率2f ωπ=得到类似的表达式。
因此,总能量检测器测定表达为：（2-4）方程式（3）和（4）可以通过傅里叶变换转换成（2-5）基线定义为δ→∞±强度,其中两个脉冲完全分离，因此，我们有：（2-6） 通过定义，干涉图可以写作：（2-7）公式（7）中2|()|E ω可表示为： （2-8）这里22|()||()|E E ωω=-，用公式（1）和关系式2()|(2/)|total I E c λπλ∝，形成因子可以表示成：（2-9）因此，干涉图包含频率谱的相干渡越辐射并可以用来推导出串长度，对于一些与高斯纵向分布：2/222/2()|()()|2|||()|2|||()|i c i c I TRE RTE e d RE RT E e d RT E d ωδωδδωωωωωωω+∞--∞+∞+∞--∞-∞∝+=+⎰⎰⎰()()i t E E t e dt ωω+∞-∞=⎰22222|||()|2|||()|RT E t dt I RT E t d ω+∞-∞∞+∞-∞⎧⎪∝⎨⎪⎩⎰⎰时域频域2*22/2||Re ()()()()2Re |||()|i c RT E t E t dt c S I I RT E e d ωδδδδωω+∞-∞∞+∞--∞⎧+⎪=-∝⎨⎪⎩⎰⎰时域频域2/21|()|()4||i c E S e d c RT ωδωδδπ+∞-∞∝⎰2/211(;)[()1]14||()i c e f n S e d N c RT NI πλδλδδπλ+∞-∞∝⨯--⎰（2-10）干涉图就变成：（2-11）高斯干涉图的FWHM是z ，因此，0.7527≈的干涉图FWHM 。
3 实验说明 目前这个试验是在飞秒加速器太赫兹SINAP 研究中心，主要由热电子射频枪，一个α磁铁，和斯坦福直线加速器中心(斯坦福线性加速器中心)类型加速管。
α的磁铁是用于产生压缩束的热离子射频枪，然后电子束是经由枪到直线加速器形成，最后由SLAC 类型管增加到20-30伏，具有高亮度的相干太赫兹辐射将发出超短束通过铝箔。
众所周知，迈克耳孙干涉仪主要取决于最大光程差的两部分相干光。
然而，如果镜子的卫面在整个扫描中保持良好对齐，如果光通过干涉仪是足够平行，那么这是唯一正确的。
然而，把平面镜移歪或摆动，因为它反应缓慢，这样就不会总是垂直入射电子束。
这使光来自反射的动镜偏离了光轴的探测器，如图1所示：图1 倾斜移动镜子使重组光束偏离迈克耳逊光学显示光轴的原理图为了计算最大允许镜倾斜，我们首先对圆形的光斑镜子引入了调制效率，可以写成： 1()2[()/]m J a a η= （3-1）在这里()m η是调制效率,1()J α是第一类贝塞尔函数,给定：22/2()z f z πσ-=22/4*()()()z S f z f z dz δσδδ+∞--∞∝+=⎰4a πσαγ= （3-2）σ是波数率(cm -1)， α是倾斜角度(弧度)，γ是光斑的半径(厘米)。
作为一般规则,一个令人满意的调制效率必须满足：()0.9m η≥（3-3） 12[()/]0.9J a a ≥ （3-4）根据科恩公式，公式（12）可约等于：（3-5） 这里222A πγ=。
假设σ= 100cm −1，r = 2.5cm ，要求容许镜倾斜保持在42.8510rad α-≤⨯,或58.7α≤弧秒。
根据上面的分析, 在整个扫描过程中必须使α小于58.7弧秒。
图2 回射的空心反光镜特性要克服迈克耳孙干涉仪倾斜的影响,我们的解决方案是用空心反光镜取代平面镜。
一个空心反光镜是一个由三个相互正交的镜子组成的装置。
对于我们的实验，空心回射器是由埃德蒙公司制造(NT46-189)；因为黄金具有良好的反射率，在太赫兹的地区范围内，所有的三镜都是金属涂层。
最重要的优点是利用空心反光镜可以将返回光沿着一条平行入射光的路径。
因此，需要精度为小于平面镜的1或2数量级，干涉仪的精度是由空心回射器本身(NT46-189)的最大宽偏差是5弧秒)和空心反光镜位置的移动决定的，空心反光镜的回射特性见图2。
4 测量结果和分析得到的长度测量干涉图显示在图3，由于分束器效率的影响，测量的干涉图的FWHM 似乎小于实际值。
因此校正就很有必要，我们通过调查获得的分束器的功率谱影响。
对于高斯束的FWHM 之间的关系，纠正后和实测的FWHM 显示在图4。
FWHM 的测量值224µm 表示高斯束长度为74µm 或248 fs 。
2212(4)14J A πσγασαπσγα≈-图3长度测量干涉图图4修正和测量FWHM高斯束之间的关系5后续工作我们开发并描述了系统产生的飞秒长度串，通过初步改善的迈克耳孙干涉仪进行了迈克耳孙干涉法对飞秒CTR电子串的测量，在半最大值在CTR干涉法下取得了长度值约248fs。
然而，在大气环境中太赫兹光被强烈吸收，这将影响测量的准确度。
根据Birch原理，虽然水吸收不会影响整体的形状谱，但将扩大串长度，这可以解释为由于水蒸气在潮湿的空气扩大分散。
因为潮湿的空气折射率不是固定在太赫兹范围内，不同的频率有不同传播速度。
因此，当它穿过空气时辐射脉冲传播，我们计算了在潮湿的空气下的真空干涉图，两种情况下的干涉图如图5所示，其中显示了增加在空气中的干涉图宽度测量，所以为了更好的精度，有必要把干涉仪放在真空中测量。
图5在真空(固体)和在潮湿的空气(虚线)的干涉图对比(E = 20MeV,Q = 0.05nC, z = 200fs)原文：Preliminary result of bunch length measurementusing a modified Michelson interferometerLIN Xu-Ling，ZHANG Jian-Bing,，LUO Feng，BEI HuaLU Shan-Liang，YU Tie-Min，DAI Zhi-MinAbstract：Based on the femtosecond accelerator device which was built at the Shanghai Institute of Applied Physics (SINAP), recently a modified far infrared Michelson interferometer has been developed to measure the length of electron bunches via the optical autocorrelation method. Compared with our former normal Michelson interferometer, we use a hollow retroreflector instead of a flat mirror as the reflective mirror. The experimental setup and results of the bunch length measurement will be described in this paper.KEY WORDS：femtosecond linear accelerator, bunch length, interferometer, hollow retroreflector1IntroductionRecent experiments on electron pulse compression have produced femtosecond electron bunches with a high peak current and brightness. Interest in short bunches arises from the requirements of high beam quality in potential applications. High quality nu- clear physics accelerators, free electron laser drive ac- celerators, next generation linear colliders, and fourth generation light sources all require short time dura- tion beam pulses[1] . Simultaneously, research into the diagnostics of the short electron bunch has played an important role in the progress. Several methods are in use or under development to measure the length of short electron bunches[2] . These generally fall into two categories: frequency-domain methods and time- domain methods. Among the time-domain methods for measuring the bunch length the use of a streak camera is well known. The streak cameras have been shown to be limited to bunch lengths longer than200fs. Additionally, streak cameras are expensive and the measurement system is sophisticated.Instead of time-domain methods, frequency-domain measurements using coherent transition radiation (CTR) from metallic foils have shown promise in the measurement of very short femtosecond electron pulses.In this paper we first present a theoretical and experimental investigation on the generation of high intensity coherent transition radiation from short electron bunches, then discuss themethod based on coherent transition radiation to measure the bunch length of femtosecond electron bunches , and then the improved experimental setup and results of the bunch length measurement are given . Finally , we analyze the e ffects of humidity in air on bunch length mea - surements and explain the plan for future investigations .2 Theoretical background2.1 Coherent transition radiationRadiation from a relativistic electron bunch such as synchrotron radiation , transition radiation , etc . intrinsically has a broad spectrum . If the wavelength of the radiation is shorter than the electron bunch length , the phases of the radiation emitted by the electrons di ffer from each another , so the radiation is incoherent . On the other hand , if the wavelength is longer than the bunch length , the radiation is coherent and the intensity of the radiation is proportional to the square of the electron numbers per bunch . The spectral intensity emitted by a bunch of N particles is given by11()()(1)()|()|tot I NI N N I f λλλλ=+- (2-1)where 1()I λ is the intensity radiated by a single electron and f (λ) is the bunch form factor [3, 4] , which is the Fourier transform of the normalized electron density distribution S (z ). For a relativistic bunch whose transverse dimension is small compared to the length , the form factor becomes()()exp[2()]f S Z i n z dz λπλ=⎰ (2-2)where n is the unit vector pointing from the center of the bunch to the observation point and z is the position vector of the electron relative to the bunch center . Obviously , a measurement of the radiation spectrum will give the form factor and the electron density distribution through the Fourier transform .2.2 Electron bunch length measurementMeasurement of electron bunch length is often done by examination of the autocorrelation of the CTR signal with a Michelson interferometer [5] . The interferometer is composed of a beam splitter , a fixed flat mirror and a movable flat mirror . The light en - tering the Michelson interferometer is split into two parts by the beam splitter . The two parts travel in two di fferent directions and are reflected back by the mirrors . After reflection the two radiation pulses are combined again and transmitted into a Golay detector to measure the light intensity .The interferogram is obtained by measuring the detector signal as a function of the path di fference in the two arms . The measured energy of the recombined radiation pulses are theradiation pulses from the fixed mirror , ()fix E TRE t =, and the radiation from the movable mirror delayed in the time /c δ, (/)move E RTE t c δ=+. Here ()R R ω= and ()T T ω= are the reflection and transmission coe fficients of the beam splitter . The intensity measured at the detector can be expressed as(2-3)where δ is the optical path di fference , c is the speed of light . Alternatively , a similar expression can be obtained in the frequency domain by adding an extra phase di fference /i c e ωδ- to the radiation from the movable arm at angular frequency 2f ωπ=. Thus , the total energy measured at the detector is expressed as(2-4) and Eqs . (3) and (4) are related by the Fourier transform(2-5)The baseline is defined as the intensity at δ → ±∞, where the two pulses are totally separated , hence , we have(2-6)By definition , the interferogram can be written as(2-7)Solving for 2|()|E ω in Eq .(7) yields(2-8)where 22|()||()|E E ωω=- . Using Eq . (1) and the relation 2()|(2/)|total I E c λπλ∝ the bunch form factor can be obtained from(2-9) 2*22()|()()|2||()()2||2|||()|I TRE t RTE t dt cRT E t E t dt RT RT E t dt c δδδ+∞-∞+∞-∞∝++=+++⎰⎰⎰2/222/2()|()()|2|||()|2|||()|i c i c I TRE RTE e d RE RT E e d RT E d ωδωδδωωωωωωω+∞--∞+∞+∞--∞-∞∝+=+⎰⎰⎰()()i t E E t e dt ωω+∞-∞=22222|||()|2|||()|RT E t dt I RT E t d ω+∞-∞∞+∞-∞⎧⎪∝⎨⎪⎩⎰⎰时域频域2*22/2||Re ()()()()2Re |||()|i c RT E t E t dt c S I I RT E e d ωδδδδωω+∞-∞∞+∞--∞⎧+⎪=-∝⎨⎪⎩⎰⎰时域频域2/21|()|()4||i c E S e d c RT ωδωδδπ+∞-∞∝⎰2/211(;)[()1]14||()i c e f n S e d N c RT NI πλδλδδπλ+∞-∞∝⨯--⎰hence , the interferogram contains the frequency spectrum of coherent transition radiation and can be used to derive the bunch length .For a bunch with Gaussian longitudinal distribution ,(2-10)the interferogram becomes(2-11)and the FWHM of this Gaussian interferogram isz . Therefore , the equivalent bunch length for a Gaussian bunch distribution is0.7527≈ times the interferogramFWHM . 3 Description of the equipmentThe present experiment was performed at the Femtosecond Accelerator in the THz Research Center of SINAP , which mainly consists of a thermionic RF gun , an α magnet , and a SLAC (Stanford Linear Accelerator Center ) type accelerating tube . The α magnet is used to compress the bunches produced by the thermionic RF gun . Then the electron beam is transported through the gun -to -linac beam line and finally accelerated up to 20—30 MeV by a SLAC type tube . Coherent THz radiation with high brightness will be emitted when super -short bunches pass through the aluminum foil [6]. It is well known that the resolution of a Michelson interferometer is mainly determined by the maximum optical path di fference of the two parts of coherent light . However , this is only true if the planes of the mirrors remain in good alignment throughout the entire scan and if the light that passes through the interferometer is su fficiently collimated . However , the moving flat mirror tends to tilt or wobble as it is retarded and , as such , will not always be perpendicular to the incident beam . This causes the light originating from the reflection of the movable mirror to deviate o ff the optical axis of the detector , as shown in Fig . 1.22/2()z f z πσ-=22/4*()()()z S f z f z dz δσδδ+∞--∞∝+=⎰Fig . 1. Schematic diagram of the Michelson optics showing how tilting the moving mirrorcauses the recombinant beams to diverge from the optical axis .In order to calculate the maximum allowable mirror tilt , we introduce first the modulation e fficiency in the case of the circular shape of the light spot on the mirrors , which could be written as1()2[()/]m J a a η= (3-1)where ()m η is the modulation e fficiency , 1()J a is the first order Bessel function , with a given by4a πσαγ= (3-2)σ is the wave number of interest (cm −1 ), α is the tilt angle (radians ) and γ is the radius of the light spot (cm ).As a general rule , a satisfactory modulation e fficiency must satisfy [7]()0.9m η≥ (3-3)12[()/]0.9J a a ≥ (3-4)According to Cohen[8] , Eq . (12) can be approximated by(3-5)where 222A πγ= . Assuming σ= 100cm −1，r = 2.5cm requires that the allowablemirror tilt be kept at a value of 42.8510rad α-≤⨯, or 58.7α≤ arc seconds . Accordingto the above analysis , α must therefore be less than 58.7 arc seconds throughout the entire scan .Fig . 2. The retroreflection property of the hollow retroreflector .To overcome the e ffect of tilt in our former Michelson interferometer , our solution is to replace the flat mirror by a hollow retroreflector . A hollow retroreflector is a device made up of three mutually orthogonal reflective mirrors . For our experiment the hollow retroreflector was made by the Edmund corporation (NT46-189); because gold has agood reflectivity in2212(4)14J A πσγασαπσγα≈-the THz region, all the three mirror are metal coated. The most important advantage of the hollow retroreflector is the fact that it can return the light along a path that is parallel to that of the incident light. As a result, the required accuracy of alignment is1or2orders of magnitude less than that of the flat mirror. The interferometric accuracy is determined by the hollow retroreflector itself(NT46-189’s maximun beam deviation is5arc seconds) and the position of the moving hollow retroreflector. A picture of the retroreflection property of the hollow retroreflector is shown in Fig. 2.4Measurement result and analysisThe interferogram obtained from a bunch length measurement is shown in Fig.3. Because of the impacts of the beam splitter efficiency, the measured FWHM of the interferogram appears to be narrower than the real value[9]. Therefore a correction became necessary, which we obtained by investigating the beam splitter effects on the power spectrum. For a Gaussian bunch the relationship between the corrected F W HM and the measured FWHM is shown in Fig.4. The measurement of224 µm FWHM indicates a Gaussian bunch length of74µm or248 fsFig. 3. Interferogram from a bunch length measurement.Fig. 4. The relationship between corrected FWHM and measured FWHMfor a Gaussian bunch.after applying corrections due to beam splitter interferences.5Future workA system to produce and characterize femtosecond length bunches has been developed and described. We carried out CTR Michelson interferometry for femtosecond electron bunch diagnostics by a preliminary improved Michelson interferometer. We achieved a bunch length evaluation of about248fs at FWHM in the CTR interferometry. However, THz light is strongly absorbed in an atmospheric environment, which will affect the accuracy of the measurement. According to Birch[10] , although water absorption will not affect the overall shape of the spectrum, it will broaden the bunch length. This broadening can be explained by dispersion due to water vapor in humid air. Because the refractive index of humid air is not constant over the THz range, different frequencies propagate with different velocities. Consequently, the radiation pulse spreads when it travels through the air. We calculated the interferograms taken in a vacuum and in humid air. Interferograms for both cases are shown in Fig. 5, displaying the increase of the interferogram width for the measurement in air. So for better precision, it is necessary to place the interferometer in a vacuum.Fig. 5. Comparison of the interferograms taken in a vacuum(solid) and in humid air(dashedline)(E = 20MeV,Q = 0.05nC, z = 200fs)谢谢下载，祝您生活愉快！。
