Factorization and Momentum-Space Resummation in Deep-Inelastic Scattering
- 格式:pdf
- 大小:482.75 KB
- 文档页数:40
密度函数理论和杜比宁方程可以用来研究活性炭纤维在多段充填过程中的吸附行为。
密度函数理论是一种分子统计力学理论,它建立在分子统计学和热力学的基础上,用来研究一种系统中分子的分布。
杜比宁方程是一种描述分子吸附行为的方程,它可以用来计算吸附层的厚度、吸附速率和吸附能量等参数。
在研究活性炭纤维多段充填过程中,可以使用密度函数理论和杜比宁方程来研究纤维表面的分子结构和吸附行为。
通过分析密度函数和杜比宁方程的解,可以得出纤维表面的分子结构以及纤维吸附的分子的种类、数量和能量。
这些信息有助于更好地理解活性炭纤维的多段充填机理。
在研究活性炭纤维的多段充填机理时,还可以使用其他理论和方法来帮助我们更好地了解这一过程。
例如,可以使用扫描电子显微镜(SEM)和透射电子显微镜(TEM)等技术来观察纤维表面的形貌和结构。
可以使用X射线衍射(XRD)和傅里叶变换红外光谱(FTIR)等技术来确定纤维表面的化学成分和结构。
还可以使用氮气吸附(BET)和旋转氧吸附(BJH)等技术来测量纤维表面的比表面积和孔结构。
通过综合运用密度函数理论、杜比宁方程和其他理论和方法,可以更全面地了解活性炭纤维的多段充填机理,从而更好地控制和优化多段充填的过程。
在研究活性炭纤维多段充填机理时,还可以使用温度敏感性测试方法来研究充填过程中纤维表面的动力学性质。
例如,可以使用动态氧吸附(DAC)或旋转杆氧吸附(ROTA)等技术来测量温度对纤维表面吸附性能的影响。
通过对比不同温度下纤维表面的吸附性能,可以更好地了解充填过程中纤维表面的动力学性质。
此外,还可以使用分子动力学模拟方法来研究纤维表面的吸附行为。
例如,可以使用拉曼光谱或红外光谱等技术来测量纤维表面的分子吸附构型。
然后,使用分子动力学模拟方法来模拟不同分子吸附构型下的纤维表面的动力学性质,帮助我们更好地了解活性炭纤维的多段充填机理。
carhart四因子模型的momentum动量因子计算英文版Calculation of Momentum Factor in the Carhart Four-Factor ModelIn the realm of financial investing, momentum is a crucial concept that has garnered significant attention from investors and academics alike. The Carhart Four-Factor Model, an extension of the popular three-factor model proposed by Eugene Fama and Kenneth R. French, incorporates momentum as a fourth factor to better explain the performance of assets.The momentum factor in the Carhart model is designed to capture the tendency of stocks with strong recent performance to continue outperforming, and vice versa for stocks with weak performance. This factor is computed by ranking stocks based on their past returns, typically over a 12-month period. Stocks with the highest returns are assigned a positive momentumscore, while stocks with the lowest returns receive a negative momentum score.To calculate the momentum factor, one needs to follow these steps:Data Collection: Gather historical return data for a universe of stocks over a specified time period, such as 12 months.Ranking: Rank the stocks based on their total returns over the chosen time period. Stocks with the highest returns will be at the top, and those with the lowest returns will be at the bottom.Momentum Scoring: Assign momentum scores to the stocks. Typically, the top-performing stocks receive a score of 1, while the bottom-performing stocks receive a score of -1. Intermediate-performing stocks can be assigned scores accordingly, such as 0.5 or -0.5, depending on their ranking.Weighting: Determine the weights for each stock based on its market capitalization or some other metric. This step ensuresthat larger stocks have a greater impact on the momentum factor.Calculation: Calculate the momentum factor by summing the weighted momentum scores of all the stocks in the universe.The momentum factor obtained from this calculation can then be used in investment strategies that aim to capitalize on the momentum effect. Investors who believe in the momentum phenomenon would allocate funds to stocks with positive momentum scores, expecting them to continue their upward trend.In conclusion, the momentum factor in the Carhart Four-Factor Model is a crucial component that helps investors identify stocks with strong momentum and potential for further gains. By incorporating momentum into their investment decisions, investors can hope to enhance their returns and mitigate the risks associated with market fluctuations.中文版Carhart四因子模型的动量因子计算在金融投资领域,动量是一个备受投资者和学者关注的重要概念。
高超声速下返回舱气动热的数值计算张钧波,张敏,刘心志南京理工大学动力工程学院,南京 (210094)E-mail :zhangm@摘 要:用非结构化网格和有限容积方法,对载人飞船返回舱在高超声速连续流下的气动热进行数值计算。
返回舱外部热通量分布采用费-里德尔(Fay-Riddell )和李斯(Lees )工程近似方程,作为对流和辐射热边界条件,求解返回舱在不同马赫数下内部温度场分布。
通过返回舱外形的优化设计,为返回舱的安全可靠和经济设计制造提供有益的参考经验。
关键词:飞船返回舱,气动热,工程算法,数值模拟1. 引言2008年4月19日,俄罗斯“联盟”号载人飞船返回舱在进入地球大气层时,由于着陆方式不当且与地面保持的角度有误,使得返回舱的天线遭焚毁,导致宇航员不能同俄地面飞行控制中心沟通。
隔热罩部分遭焚毁和舱门严重损坏,俄罗斯和韩国宇航员不得不从返回舱挣脱,并且有严重灼伤。
同时“联盟”号着陆后,引起附近灌木起火。
气动热是载人飞船返回舱的关键问题之一,以7.91 km/s 第一宇宙速度飞行的返回舱单位动能为31.3 MJ/m 2,如此巨大的能量如果都转化成热能,就会把载人飞船返回舱的材料烧坏。
所以必须做好返回舱的气动力,气动热和热防护的设计,尽量把该巨大的能量扩散到防护材料和周围介质中去,以保证返回舱不被烧毁和舱内正常的工作和生存环境[1-3]。
目前世界上小升阻比载人飞船返回舱,如星座号、阿波罗号和联盟号,以及我国的神州号等都无一例外的选择了球冠倒锥形。
返回舱采用这种形式,能使锥面热流率大为减小。
因为钝头体热流密度与曲率半径成反比,曲率半径越大,热流率越小。
这不但在气动力上是优秀的,而且也是气动热和热防护上所需要的[1]。
本文在采用球冠倒锥形的基础上,采用作者开发的软件数值计算返回舱气动热的相关问题。
2. 载人返回舱几何外形设计沿y 轴的方向,我们把返回舱(见图1所示)分成三个区域,母线方程如下:底部球冠圆的母线方程:222( 1.3)( 1.9) 2.3125x y −+−=倒锥母线的直线段方程: 3.7293.7299.7y x y x ==−+ 倒锥母线圆弧段方程:222( 1.3)(0.9)x y r ++−=其中:L=1800 mm, a=100mm ,b=1000mm ,C=400mm ,d=1140mm ,R=1300mm ,r=940mm ,∠p=150,∠q=500。
第6卷 第3期2013年6月 中国光学 Chinese Optics Vol.6 No.3June 2013 收稿日期:2013⁃02⁃17;修订日期:2013⁃04⁃15 基金项目:国家自然科学基金资助项目(No.10834015;No.61077082);陕西省科技新星资助项目(No.2012KJXX⁃27);陕西省光电技术与功能材料省部共建国家重点实验室培育基地基金资助项目(No.ZS12018)文章编号 1674⁃2915(2013)03⁃0283⁃14太赫兹波段超材料的制作、设计及应用潘学聪1,姚泽瀚2,徐新龙1,2∗,汪 力1(1.中国科学院物理研究所北京凝聚态物理国家实验室,北京100190;2.西北大学光子学与光子技术研究所光电技术与功能材料国家重点实验室培育基地,陕西西安710069)摘要:本文从制作方法、结构设计和材料选择几方面综述了超材料在太赫兹波段的电磁响应特性和潜在应用。
首先,介绍了获得不同维度、具有特异电磁响应以及结构可调超材料的各种微加工制作方法,进而分析和讨论了超材料的电磁响应特性。
文中指出,结构设计可以控制超材料的电磁响应特性,如各向异性、双各向异性、偏振调制、多频响应、宽带响应、不对称透射、旋光性和超吸收等。
超材料的电磁响应依赖于周围微环境的介电性质,因而可用于制作对环境敏感的传感器件。
此外,电光、磁光、相变、温度敏感等功能材料的引入可以获得光场、电场、磁场、温度等主动控制的太赫兹功能器件。
最后,简单介绍了超材料在太赫兹波段进一步发展所面临的机遇和挑战。
关 键 词:超材料;太赫兹技术;结构设计;调制;偏振中图分类号:O441;TB34 文献标识码:A doi:10.3788/CO.20130603.0283Fabrication ,design and application of THz metamaterialsPAN Xue⁃cong 1,YAO Ze⁃han 2,XU Xin⁃long 1,2∗,WANG Li 1(1.Beijing National Laboratory for Condensed Matter Physics ,Institute of Physics ,Chinese Academy of Sciences ,Beijing 100190,China ;2.State Key Laboratory Incubation Base of Photoelectric Technology and Functional Materials ,Institute of Photonics &Photon⁃Technology ,Northwest University ,Xi′an 710069,China )∗Corresponding author ,E⁃mail :xlxuphy@ Abstract :In this paper,the electromagnetic responses and potential applications of THz metamaterials are re⁃viewed through the focus on fabrication,unit structure design,and material selection,respectively.It de⁃scribes different kinds of fabrication technologies for obtaining metamaterials with special electromagnetic re⁃sponses such as magnetic resonance and reconfigurable tunability,which is helpful for further understanding of electromagnetic resonances in metamaterials.The paper analyzes the electromagnetic response characteristics in detail and points out that the unit structure design can be used to obtain desired electromagnetic characteris⁃tics,such as anisotropy,bianisotropy,polarization modulation,multiband response,broadband response,asymmetric transmission,optical activity,and perfect absorption,etc .The dependence of electromagnetic re⁃sponses upon surrounding dielectrics can be used not only to control resonant frequency by a proper substrateselection,but also for sensing applications.Furthermore,the introduction of functional materials with control⁃lable dielectric properties by external optical field,electrical field,magnetic field and temperature has the po⁃tential to achieve tunable metamaterials,which is highly desirable for THz functional devices.Finally,the op⁃portunities and challenges for further developments of THz metamaterials are briefly introduced.Key words:metamaterials;THz technology;structure design;modulation;polarization1 引 言 通过对自然材料的裁剪、加工和设计,从而实现对电子、光子以及其他一些元激发准粒子的人为调控,一直是光电科学研究的重点。
相空间重构方法用于动力学的分析研究一、引言动力学是研究物体运动过程的科学,是物理学、力学等学科的重要分支。
为了深入理解物体的运动状态和规律,研究人员提出了许多分析方法。
其中,相空间重构方法是一种重要的手段,通过将动力学系统的状态空间转化为相空间,能够更直观地揭示运动的特征和行为。
本文将介绍相空间重构方法的原理和应用,以及在动力学研究中的实际应用。
二、相空间重构方法的原理相空间重构方法是由统计力学中的一系列概念发展而来的。
在动力学系统中,物体的状态可以用一组状态变量来描述,比如位置、速度等。
在经典动力学中,这些状态变量是连续变量。
为了方便分析,我们需要将连续的状态空间转化为离散的相空间。
相空间重构方法的基本思想是:从动力学系统中选择合适的状态变量,并采样它们的值,然后按照一定的规则将这些值组合起来,形成一个表示系统状态的向量。
通过收集足够多的这样的向量,就可以构建出一个描述系统状态的相空间。
为了保证相空间的有效性,我们通常需要选择合适的状态变量,并在采样时遵循一定的原则,比如均匀采样、等时间间隔采样等。
三、相空间重构方法的应用相空间重构方法在动力学研究中有着广泛的应用。
下面将介绍几个典型的应用领域。
1. 混沌系统分析混沌系统是指在非线性动力学中表现出复杂非周期性行为的系统。
相空间重构方法可以帮助我们理解和描述混沌现象。
通过将混沌系统的状态空间重构为相空间,我们可以观察到混沌轨迹的特征,比如奇异吸引子、分岔等。
这对于深入研究混沌系统的行为规律非常重要。
2. 动力学预测相空间重构方法在动力学预测中也起到了关键作用。
通过在相空间中查找轨迹的性质和规律,可以预测系统的未来行为。
这对于一些需要提前做出决策的应用场景非常重要,比如天气预测、股票走势预测等。
3. 物理系统建模相空间重构方法还可以用于物理系统的建模。
通过对物理系统进行观测和采样,然后进行相空间重构,我们可以得到一个描述物理系统状态的数学模型。
这些模型可以用于进一步研究系统的性质和行为,比如流体力学中的流动现象研究、量子力学中的概率分布研究等。
《多功能微纳米过渡金属羰基CO释放分子(CORMs)复合体系的构建与性能研究》篇一一、引言随着科技的发展和纳米科技的兴起,对于材料的多功能性及高效性需求愈发显著。
微纳米材料中的多功能过渡金属羰基CO 释放分子(CORMs)因其独特的光学、电子和催化性质在许多领域如医药、环保和能源领域都有重要的应用。
因此,本篇论文着重探讨了多功能微纳米过渡金属羰基CO释放分子(CORMs)复合体系的构建以及其性能研究。
二、CORMs及其复合体系的构建2.1 CORMs的介绍CORMs是一种以过渡金属为基础的有机化合物,它们可以控制地释放CO气体,这使得它们在多个领域具有独特的应用。
其核心结构包括过渡金属原子与CO的键合。
2.2 微纳米过渡金属CORMs的构建微纳米尺寸的CORMs,因其更小的尺寸和更大的比表面积,使得它们在反应中具有更高的活性和效率。
我们通过特定的合成方法,成功构建了微纳米过渡金属CORMs。
2.3 复合体系的构建为了进一步增强CORMs的性能,我们通过与其他材料进行复合,构建了多功能微纳米过渡金属CORMs复合体系。
这些复合体系不仅可以增强CORMs的稳定性,同时也能提升其反应活性和选择性。
三、性能研究3.1 光学性能研究通过紫外-可见光谱分析,我们发现微纳米CORMs在特定波长下具有明显的吸收峰,这表明它们具有独特的光学性质。
同时,复合体系的光学性能也得到了显著提升。
3.2 电子性能研究利用电子显微镜和电子能谱分析,我们发现微纳米CORMs 具有较高的电子传输效率。
同时,复合体系中的电子传输速度也得到了显著提升。
3.3 催化性能研究我们通过一系列的催化实验发现,微纳米CORMs复合体系在多种反应中表现出良好的催化活性。
特别是在某些有机合成反应中,其催化效率远高于传统的催化剂。
四、结论本论文研究了多功能微纳米过渡金属羰基CO释放分子(CORMs)复合体系的构建与性能。
通过实验和理论分析,我们发现这种复合体系在光学、电子和催化性能上均表现出良好的表现。
chen混沌系统方程解释说明1. 引言1.1 概述混沌系统是指具有不可预测性和高度敏感依赖于初始条件的动力学系统。
这些系统在数学上表现出复杂的、非周期的行为,其演化过程无法由常规的微分方程描述。
Chen混沌系统是其中一种经典的混沌系统模型,由Chen等人在20世纪90年代提出,并引起了广泛关注。
1.2 文章结构本文将首先介绍混沌系统方程的背景知识,包括其理论基础、历史发展和应用领域。
接着详细解释Chen混沌系统方程的定义和属性,并探讨其数学表达式、相空间描述以及Lyapunov指数和混沌性质。
随后,我们将对Chen混沌系统方程进行动力学行为分析和模拟探究,包括平衡点和稳定性分析、流场特征与相轨迹演化以及参数选择与动力学行为模拟。
最后,文章将总结对Chen混沌系统方程的研究成果,并展望未来研究的方向与挑战。
1.3 目的本文旨在对Chen混沌系统方程进行全面的解释和说明。
通过详细介绍Chen混沌系统方程的数学表达式、属性特征以及动力学行为分析,读者能够对该混沌系统模型有更深入的理解。
此外,本文还将探讨未来研究该方程可能面临的挑战和可行的研究方向,为相关领域的学者提供参考和启示。
2. 混沌系统方程的背景2.1 理论基础混沌系统是一类具有无规则行为和高度敏感依赖于初始条件的动力学系统。
与传统的线性系统不同,混沌系统表现出不可预测性和复杂性,其运动轨迹在相空间中呈非周期性而且高度复杂。
正是这种无规律的行为给混沌系统带来了很多新奇的特性和应用。
混沌理论的发展起源于随机过程和动力学领域,早期由著名数学家洛伦茨所提出的洛伦兹吸引子模型成为了研究混沌现象的重要基础。
此后,多个混沌模型被提出并广泛研究,其中包括经典的Henon映射、Logistic映射以及Chua电路等。
2.2 历史发展Chen混沌系统方程是由陈氏夫妇于1999年提出的一种三维非线性动力学方程。
这个方程通过调节参数可以实现从周期运动到混沌现象的转变,在控制理论、信息加密等领域得到了广泛应用。
我所知道的计算流体力学大牛们计算流体力学(CFD)是一门涉及数学、物理学和计算机科学的交叉学科,广泛应用于工程和科学领域。
下面是一些在CFD领域做出重要贡献的大牛们。
1. Jameson, Antony:Jameson教授是CFD领域的先驱之一,他提出了著名的领先-追赶格式方法,用于解决非稳态问题。
他还开发了自适应网格技术,用于提高CFD计算的精度和效率。
2. Anderson, John D. Jr.:Anderson教授是航空航天工程和计算流体力学领域的一位重要学者。
他在CFD中的突出贡献包括开发了NASA 的三维多块结构网格生成软件。
他还合著了《计算流体力学和有限元法》等经典著作。
3. Roache, Patrick J.:Roache教授被誉为CFD方法验证和验证的权威。
他提出了CFD验证的概念,并开发了一系列方法和标准,以确保CFD计算的准确性和可信度。
他的贡献促进了CFD方法的发展和应用。
4. Chorin, Alexandre Joel:Chorin教授提出了著名的分数阶步进方法(Fractional Step Method),用于解决不可压缩流动的Navier-Stokes方程。
这种方法极大地简化了CFD计算的过程,并被广泛应用于工程实践中。
5. Melnik, Roderick V. N.:Melnik教授在CFD领域做出了重要贡献,特别是在多孔介质流动和多孔介质材料的数值模拟方面。
他开发了多孔介质流动的理论模型和数值方法,并应用于地球物理、生物医学和工程等领域。
6. Ferziger, Joel H.:Ferziger教授是CFD领域的重要学者之一,他合著了《计算流体力学》这本广泛使用的教材。
他的研究涵盖了流体力学的各个方面,包括湍流模拟、边界条件处理和数值方法的改进等。
7. Peric, Milovan:Peric教授在CFD的边界条件处理和有限体积方法方面做出了重要贡献。
a r X i v :h e p -p h /0607228v 1 20 J u l 2006CLNS 06/1971FERMILAB-PUB-06-242-TSFB/CPP-06-32SI-HEP-2006-09July 20,2006Factorization and Momentum-Space Resummation inDeep-Inelastic ScatteringThomas Becher a ,Matthias Neubert b,c ,and Ben D.Pecjak d a Fermi National Accelerator Laboratory P.O.Box 500,Batavia,IL 60510,U.S.A.b Institute for High-Energy Phenomenology Newman Laboratory for Elementary-Particle Physics,Cornell University Ithaca,NY 14853,U.S.A.c Institut f¨u r Physik (ThEP),Johannes Gutenberg-Universit¨a t D–55099Mainz,Germany d Theoretische Physik 1,Fachbereich Physik,Universit¨a t Siegen D–57068Siegen,Germany Abstract Renormalization-group methods in soft-collinear effective theory are used to perform the resummation of large perturbative logarithms for deep-inelastic scattering in the thresh-old region x →1.The factorization theorem for the structure function F 2(x,Q 2)for x →1is rederived in the effective theory,whereby contributions from the hard scale Q 2and the jet scale Q 2(1−x )are encoded in Wilson coefficients of effective-theory oper-ators.Resummation is achieved by solving the evolution equations for these operators.Simple analytic results for the resummed expressions are obtained directly in momentum space,and are free of the Landau-pole singularities inherent to the traditional moment-space results.We show analytically that the two methods are nonetheless equivalent order by order in the perturbative expansion,and perform a numerical comparison up to next-to-next-to-leading order in renormalization-group improved perturbation theory.1IntroductionIt is well known thatfixed-order perturbation theory is not reliable for quantities involving several disparate scales.In such cases,higher-order corrections are enhanced by large loga-rithms of scale ratios.The standard solution to this problem is to split the calculation into a series of single-scale problems by successively integrating out the physics associated with the largest remaining scale.Perturbative logarithms are then resummed by renormalization-group (RG)evolution from the larger scales to the smaller ones.For collider processes,resummation is traditionally performed by other means,since it was not always clear how to systematically integrate out the physics associated with high scales in such cases.The simplest example of a high-energy process with a scale hierarchy which necessitates resummation is deep-inelastic scattering(DIS)in the threshold region.As the Bjorken scaling variable x→1,the invariant mass of the hadronic system produced in the decay,M X= Q x(neglecting the nucleon mass),becomes much smaller than the momentum transfer Q.The presence of the two scales is manifest in the QCD factorization theorem[1,2,3]Fns2(x,Q2)=H(Q2,µ)Q2 1x dz z,µ x z,µ ,(1)for the non-singlet part of the structure function F2(x,Q2).The result(1)is valid in the threshold region at leading power in M2X/Q2≈(1−x)andΛ2QCD/M2X.As long as M X≫ΛQCD,both the jet function J(M2X,µ)and the hard function H(Q2,µ)can be evaluated in perturbation theory,whereas the parton distribution functionφns q(ξ,µ)is a non-perturbative object.The result for the hard function involves single and(Sudakov)double logarithms of the formαn s ln m(Q/µ),with m≤2n,while the integral over the jet function produces logarithms αn s ln m(M X/µ).Irrespective of the value of the renormalization scaleµ,thefixed-order result contains large logarithms.Traditionally,the resummation of these logarithms is performed in moment space.The threshold region of small M X is probed by large-N moments.The relevant scale in Mellin space is Q/√1−z(1−z)Q2µ2dk21−z) .(2)The functions A q and B q are determined by matching with results fromfixed-order perturba-tion theory and are currently known at three-loop order,enabling a nearly complete threshold resummation to next-to-next-to-next-to-leading logarithmic(N3LL)accuracy[4].The re-summed momentum-space structure function F2(x,Q2)is obtained from the moment-space expression by an inverse Mellin transformation.This approach to threshold resummation has several drawbacks.Thefirst is related to integrations over the Landau pole in the running coupling.These occur twice:once in the integrals over the functions A q and B q in the resummation exponent,and once again when the inverse Mellin transform is taken to obtain results in momentum space.To perform theresummation one needs to specify a prescription for how to deal with these poles.Various methods have been proposed in the literature,such as the“minimal prescription”[5]or the “tower expansion”[6].The difference between these prescriptions is a power-suppressed effect. Since factorization theorems do receive power corrections,this does not appear as a problem atfirst sight.However,as discussed in[7],the Landau-pole singularity in the resummed expression can induce large unphysical power corrections.In the example of the Drell-Yan process,the ambiguity in the threshold resummation amounts to a power correction of order ΛQCD/M X,while the physical power corrections to the process scale asΛ2QCD/M2X.The fact that resummations with RG methods[7,8]do not involve Landau-pole ambiguities illustrates that these effects do not have a direct physical interpretation.In particular,a Landau-pole ambiguity does not necessarily imply the presence of a commensurate renormalon ambiguity [7].Further drawbacks are that in the traditional resummation formalism the separation of contributions from the hard and jet scales is not transparent,and while the function A q has a clear interpretation as the cusp anomalous dimension familiar from the renormalization theory of Wilson lines[9,10],the function B q is not easily identified with afield-theoretical object.In this paper we use RG techniques to perform the resummation of perturbative loga-rithms directly in momentum space.The starting point is the factorization formula(1),which we rederive using soft-collinear effective theory(SCET)[11,12,13,14].In this framework, the hard function H and the jet function J are matching coefficients.The hard function arises from afirst matching step,in which the electroweak current is matched onto a corre-sponding effective-theory current operator.In a second step,the partons associated with the hadronicfinal state are integrated out,giving rise to the jet function.Threshold logarithms are resummed by solving the RG equations for these matching functions,using techniques presented in[15].Existing results from higher-order perturbative calculations enable us to perform the matching and resummation up to next-to-next-to-leading order(NNLO)in RG-improved perturbation theory,corresponding to the N3LL approximation in the standard approach.We show that the results obtained in momentum space are formally equivalent to the more familiar moment-space results by deriving a formula which connects them order by order in perturbation theory.However,integrals over the Landau pole never appear in our momentum-space formulation,and the effective-theory matching functions and anoma-lous dimensions have a clearfield-theoretical interpretation.Furthermore,we obtain a simple analytic expression for the resummed structure function,while the Mellin inversion which is necessary in the traditional approach can only be performed numerically.As a result,it is straightforward to match our resummed expressions ontofixed-order calculations valid outside the threshold region.Finally,we stress that our approach to resummation in x-space is free of the pathologies related to large unphysical power ambiguities found in[5].In fact,it exhibits a better apparent perturbative convergence than the conventional approach.In the context of SCET,the generic factorization formula for DIS has been discussed previously in[16],while the case x→1has been studied in[17,18,19,20,21,22].These papers make conflicting statements about the factorization properties of DIS in the endpoint region.Most of the differences are resolved after observing that,near the endpoint,the parton distribution function receives contributions from two distinct non-perturbative modes. While the two modes cannot be factorized perturbatively,their presence must nonetheless be taken into account to correctly translate the effective-theory result into the QCD factorizationtheorem(1).In[7,17,19,21,23],the resummation was performed by solving the RG equations for the moments.This avoids Landau-pole singularities in the exponent,but as we show here, it is possible to solve the equations directly in momentum space.While threshold resummation in DIS is of limited phenomenological importance,it is a relatively simple process for which the perturbative results needed in our calculation are known at NNLO.For this reason,it provides an especially instructive example with which to develop our resummation formalism.However,many other processes fulfill factorization theorems of the same structure,in which the rate factorizes into a hard contribution times a jet function convoluted with a nonperturbative matrix element,and our formalism also applies to these cases.An example is heavy-particle production near threshold,which includes the Drell-Yan process in the limit where the invariant mass of the produced lepton pair is close to the center-of-mass energy in the collision,as well as Higgs production in the same kinematic region.Another example is provided by inclusive B-meson decays in the endpoint region.Our final result for the resummed DIS structure function is very similar to the factorized expression for radiative B→X sγdecay as derived in[24,25].In fact,both processes involve the same jet function,given by the quark propagator in light-cone gauge.The outline of the paper is as follows.In Section2we use SCET to obtain the QCD factor-ization formula for DIS near the endpoint,providing a translation between the effective theory and standard discussions.In Section3we work out the technique for threshold resummation in momentum space,derive a compact expression for the factorized structure function,and give results for the perturbative matching coefficients valid to NNLO in perturbation theory. In Section4we convert our results to moment space and show how they are connected to those obtained in the standard approach.The Appendix gives the perturbative expansions of the RG functions used in our analysis.2Factorization in DISIn this section we derive the QCD factorization formula for the non-singlet DIS structure function F ns2(x,Q2),using the technology of SCET[11,12,13,14].We consider DIS of electrons offa nuclear target,e−+N(p)→e−+X(P),as illustrated in Figure1.All non-trivial hadronic physics is encoded in the hadronic tensorWµν(p,q)=i d4x e iq·x N(p)|T{J†µ(x)Jν(0)}|N(p)= qµqνq2 pν−qνp·q,(4)2p·qwhere q=P−p is the momentum of the virtual photon,and x is the Bjorken scaling variable.For simplicity we focus on theflavor non-singlet component of the cross section,2mxQ2±1.(6)We shall discuss the different regions in the Breit frame,where thefinal-state hadronic jet moves along the z-direction,while the target nucleon moves in the opposite direction.The light-cone projections of the relevant momenta are(all perpendicular components vanish by choice of coordinates)n·q=−Q,¯n·q=Q,n·p=m eη=QQ+...,¯n·p=m e−η=m2xx +m2xQ+...,(7)where the neglected terms are of order m4x3/Q3.The invariant mass M X of thefinal-state hadronic jet is given byM2X=P2=Q21−xx≫m2.(8)In the last step we have used that for an inclusive process the jet mass must be much larger than m∼ΛQCD.Otherwise,the cross section cannot be analyzed using short-distance methods.While the momentum of the virtual photon isfixed by kinematics,thefinal-state jet and target nucleon consist of jets of near on-shell partons,whose momenta have scalings consistent with the relations above.We introduce a small expansion parameterλ∼m/Q∼ΛQCD/Q and quote the components(p+,p−,p⊥)of parton momenta in units of Q.Assumefirst that x=O(1)is not very close to1.Then,for the purposes of power counting,it follows that q∼Q(1,1,0),p∼Q(1,λ2,0),and P∼Q(1,1,0).The partons making up the initial and final hadronic states have generic scalingstarget nucleon:p¯c∼Q(1,λ2,λ)(anti-collinear),final-state jet:p h∼Q(1,1,1)(hard),(9) where the term“anti-collinear”refers to collinearfields propagating in the negative z-direction. These relations change for the special case where x is close to1,such thatǫ=1−x becomes parametrically small.The momentum of thefinal-state jet now scales like P∼Q(ǫ,1,0). While the valence quark in the target nucleon struck by the photon still carries an anti-collinear momentum p valence∼Q(1,λ2,λ),the remaining partons in the nucleon now have total momentum scaling like p−p valence∼Q(ǫ,λ2,λ).Consequently,the partons making up thefinal-state jet and the target remnant jet have momenta scaling likefinal-state jet:p hc∼Q(ǫ,1,√ǫλ)(soft-collinear).(10)In these relations,the scaling of the perpendicular momentum components follows from the requirement that the individual partons be nearly on-shell.The terminology for the“hard-collinear”and“soft-collinear”modes follows[26]and[27].In the traditional literature on factorization in DIS[1,2,3]the soft-collinear modes were referred to as“soft”.Relation(8) implies thatλ2≪ǫ≪1,and there is no need to specify the relative scaling betweenǫandλin more detail.All that matters for the factorization analysis is that p2hc∼Q2ǫis a perturbative scale,while p2¯c∼Q2λ2is not.The discussion of factorization for the generic case,where x=O(1)but not very close to1,is straightforward[16].Hard modes are described by QCD,whereas the anti-collinear partons making up the target nucleon can be described in SCET.There is no need to include any other modes,since the only relevant regions are hard and anti-collinear.In interactions of the anti-collinearfields with hardfields,only the large plus components p¯c+∼Q of the anti-collinear momenta should be kept at leading order in power counting.Correspondingly, the anti-collinearfields must be multipole expanded about x−=(¯n·x)n/2,i.e.,φ¯c(x)=φ¯c(x−)+....Integrating out the hard modes by matching onto SCET yields an expression for the discontinuity of the hadronic tensor of the form1xC(Q2,x/ξ,µ) dt2γνψ(0)|N(p) ,(11)where C=δ(1−x/ξ)+O(αs)is a matching coefficient in the effective theory,and the object [tn,0]is a straight Wilson line along the n light-cone.We have used that the SCET Lagrangian for a single collinear sector is equivalent to the QCD Lagrangian[12]in order to replace the SCETfields by the usual QCDfields.The identification of the nucleon matrix element with the QCD parton distribution function is then automatic(see relation(23)below),and one arrives at the standard factorization formula.The derivation of the factorization formula for x→1is more complicated.It involves a two-step matching procedure similar to that used for inclusive semi-leptonic and radiative B decays in the endpoint region[24,25,26,28].In afirst matching step,hard modes are integrated out by matching QCD onto a version of SCET containing hard-collinear,anti-collinear,and soft-collinearfields.We will refer to this intermediate effective theory as SCET(hc,¯c,sc)for short. The matching function associated with thisfirst step is the hard coefficient C V.Because the sum of a hard-collinear momentum and an anti-collinear momentum has an invariant mass (p hc+p¯c)2∼Q2and must be counted as hard,the intermediate effective Lagrangian does not contain vertices coupling the hard-collinearfields to anti-collinear ones.Thesefields interact only through the exchange of soft-collinear“messenger”fields.However,the soft-collinear modes can be decoupled from the hard-collinear ones by means of afield redefinition.After this decoupling,it is possible to integrate out the hard-collinear scale by matching onto a low-energy theory SCET(¯c,sc)involving only anti-collinear and soft-collinear modes.The matching function associated with this step is the jet function J.Having integrated out the perturbative modes,thefinal step is to evaluate the matrix element of the remaining operator defined in the low-energy effective theory.An important part of the factorization analysis is to show that this matrix element is equivalent to the QCD parton distribution function evaluated in the limit x→1,as studied e.g.in[1,3].We will show that this is indeed the case,and that the soft-collinear modes play an important role in this identification.The appropriate Lagrangian for SCET(hc,¯c,sc)is a generalization of the effective La-grangian for collinear and soft-collinearfields derived in[27,29].It contains hard-collinear quark and gluonfieldsξhc and A hc,anti-collinear quark and gluonfieldsξ¯c and A¯c,and soft-collinear quark and gluonfieldsθsc and A sc.The hard-collinearfields move along the z-direction,and hence/nξhc=0.The anti-collinear and soft-collinearfields move in the opposite direction,so/¯nξ¯c=0and/¯nθsc=0.The two collinear sectors can only interact via soft-collinear exchange,and at leading power only soft-collinear gluons are involved in these interactions.The corresponding effective Lagrangian at leading order in the expansion parametersǫandλis[27,29]L SCET(y)=¯ξhc/¯n212[i¯n·D¯c+g¯n·A sc(y+)]ξ¯c−¯ξ¯c i/D¯c⊥/n in·D¯ci/D¯c⊥ξ¯c+pure glue terms+soft-collinear Lagrangian,(12) where allfields without position argument are to be evaluated at the point y.The effective Lagrangian is invariant under a set of hard-collinear,anti-collinear,and soft-collinear gauge transformations,whose precise form can be found in[27,30].An important property of the SCET Lagrangian is that soft-collinear gluons can be de-coupled from the hard-collinear and anti-collinearfields throughfield redefinitions involving Wilson lines[12,27].This decoupling is essential for the factorization analysis below.Di-agrammatic factorization proofs also rely on the decoupling of“soft”gluons from collinear fields.The underlying physics is that soft gluons couple to collinear partons through eikonal vertices,a feature explicit in the SCET Lagrangian(12).2.1Matching of the currentThefirst step in the factorization procedure is to integrate out hardfluctuations by match-ing QCD onto the intermediate effective theory SCET(hc,¯c,sc).The kinematic restrictions implied by the limit x→1simplify thisfirst matching step.Since we are dealing with the region of phase space where thefinal-state jet is hard-collinear,there are no contributions to the hadronic tensor where the anti-collinear partons at points0and x are connected by hard gluons.It is therefore sufficient to integrate out hardfluctuations at the level of the elec-tromagnetic current.Time-ordered products of two currents are not needed until the second step.We match the QCD current Jµ(x)=(¯ψγµψ)(x)onto a current in SCET containing a hard-collinear quark and an anti-collinear anti-quark.The form of the resulting operator is dictated by gauge invariance.The appropriate matching relations for the QCDfields areψhc(x)→(W†hcξhc)(x),ψ¯c(x)→(W†¯cξ¯c)(x−),(13) where W hc is the hard-collinear Wilson lineW hc(x)=P exp ig 0−∞ds¯n·A hc(x+s¯n) (14)along the¯n-direction,and W¯c is the analogous anti-collinear Wilson line along the n-direction. The multipole expansion in(13)requires some explanation.In the hadronic tensor(3),the points0and x are connected by a hard-collinear jet propagating through a cloud of soft-collinear partons.This implies that the position argument x scales as a hard-collinear quantity, x∼(1,ǫ−1,ǫ−1multipole expansion of Lagrangian interactions between soft-collinear and anti-collinearfields [27].Therefore,when matching the current operator onto SCET,one must multipole expand both the anti-collinear and soft-collinearfields about x−.The two expressions in(13)are invariant under hard-collinear and anti-collinear gauge transformations,while under a soft-collinear gauge transformation both compositefields, W†hcξhc and W†¯cξ¯c,transform into U sc(x−)times themselves.Thus,at tree level the gauge-invariant matching relation for the current is(¯ψγµψ)(x)→(¯ξ¯c W¯c)(x−)γµ⊥(W†hcξhc)(x).(15) Only a single Dirac structure is possible for massless quarks.Beyond tree level the matching relation at leading power gets generalized to(see the analogous discussions in[13,29])(¯ψγµψ)(x)→ dt C V(t,n·q,µ)(¯ξ¯c W¯c)(x−)γµ⊥(W†hcξhc)(x+t¯n)=C V(−n·q¯n·P,µ)(¯ξ¯c W¯c)(x−)γµ⊥(W†hcξhc)(x).(16) In thefirst line we have used that¯n·∂derivatives of hard-collinearfields are unsuppressed in SCET power counting,allowing for arbitrary displacements of thesefields along the¯n light-cone.In the second line,the object P is the hard-collinear momentum operator,and the Wilson coefficient C V is the Fourier transform of the position-space Wilson coefficient C V appearing in thefirst line.In the case at hand,the relevant components−n·q≈¯n·P≈Q arefixed by kinematics(see the relations(7)),and so we may write C V(Q2,µ)for simplicity.2.2Matching of the hadronic tensorThe next step in the matching procedure is to evaluate the hadronic tensor in the intermediate effective theory.Inserting the SCET current(16)into(3),wefind the leading-power expressionWµν(p,q)→|C V(Q2,µ)|2i d4x e iq·x(17)× N(p)|T{(¯ξ¯c W¯c)(x−)γµ⊥(W†hcξhc)(x)(¯ξhc W hc)(0)γν⊥(W†¯cξ¯c)(0)}|N(p) .The interactions of soft-collinear gluons with hard-collinearfields in(12)can be removed by thefield redefinitions[12,27](x)S†n(x−),(18)ξhc(x)→S n(x−)ξ(0)hc(x),Aµhc(x)→S n(x−)Aµ(0)hcwhich imply(W†hcξhc)(x)→S n(x−)(W(0)†ξ(0)hc)(x).HerehcS n(x)=P exp ig 0−∞ds n·A sc(x+sn) (19)is a soft-collinear Wilson line along the n-direction.The redefined hard-collinearfields with superscripts“(0)”are decoupled from soft-collinearfields and thus interact only among them-selves.After thefield redefinition the hadronic matrix element in(17)factorizes into a vacuum(a)(b)(c)Figure2:Examples of diagrams involving anti-collinear gluon exchange.The dashed(dot-ted)lines represent anti-collinear(hard-collinear)quark lines.The wavy lines represent the electromagnetic currents.In graph(a)the anti-collinear gluon is part of the initial-state(nu-cleon)jet and thefinal-state propagator is hard-collinear,as required in the effective theory.In graphs(b)and(c)the anti-collinear gluon is part of thefinal-state jet,whose invariant mass then becomes hard.Graphs(b)and(c)are therefore not part of the effective-theoryrepresentation of the hadronic tensor as x→1.matrix element of hard-collinearfields and a nucleon matrix element of anti-collinear and soft-collinearfields.In the second matching step,we“integrate out”the hard-collinearfields,which can be done using perturbation theory because the hard-collinear scale is a short-distance scale,p2hc∼Q2(1−x)≫Λ2QCD.Since in a single(hard-)collinear sector SCET is equivalent to full QCD[13],the vacuum matrix element of hard-collinearfields can be rewritten in terms of theQCD matrix element[31]0|T{(W(0)†hcξ(0)hc)(x)(¯ξ(0)hc W(0)hc)(0)}|0 = 0|T /n/¯n4 |0= d4k2¯n·k J(k2,µ).(20)The object W(x)denotes a Wilson line analogous to(14)but with gaugefields in full QCD. Color indices are suppressed;the correlator is proportional to the unit matrix in color space. We define the jet function through the imaginary part of J as(see e.g.[32])J(p2,µ)=1if the SCET Feynman rules used for the matching of the electromagnetic current are naively applied to the hadronic tensor.We can construct a set of Feynman rules appropriate for the hadronic tensor by introducing different anti-collinear fields for the “in”and “out”states in the forward-scattering amplitude,and restricting interactions between the two anti-collinear sectors to soft-collinear exchange.These effective-theory Feynman rules produce graphs such as that in Figure 2(a),but not those in Figure 2(b)and (c).For simplicity of notation,we will suppress the “in”and “out”labels on the anti-collinear fields,but one must make this distinction when evaluating the hadronic tensor in the effective theory.After integrating out the hard-collinear fields,the resulting nucleon matrix element in the low-energy effective theory can be reduced toN (p )|(¯ξ¯c W ¯c )(x −)S n (x −)γµ⊥/n 2(W †¯c ξ¯c )(0)|N (p ) ,(22)where [x −,0]sc =S n (x −)S †n (0)is a straight Wilson line of soft-collinear gluon fields along the n light-cone.In the second line we have defined the objects g µν⊥=g µν−12ǫµναβ¯n αn β,and also used that /¯n ξ¯c =0.The anti-symmetric structure vanishes after averaging over the nucleon spin.The appearance of the symmetric structure g µν⊥impliesthe Callan-Gross relation Q 2W 2=4x 2W 1at leading power and to all orders in perturbation theory.Hereafter,we thus focus on the structure function W 1.Consider now the standard definition of the quark distribution function in QCD [33],φns q (ξ,µ)=12ψ(0)|N (p ) ,(23)where [tn,0]is a straight Wilson line of gauge fields in full QCD,and the superscript “ns”indicates the flavor non-singlet component of the distribution function.In the Breit frame,where the proton moves along the ¯n -direction,ψand ¯ψcan be considered anti-collinear fields.For generic values of ξthese fields carry only a portion of the proton’s longitudinal momentum.The remaining portion (1−ξ)n ·p is still large and can be shared between other anti-collinear partons exchanged between the two points 0and tn .A different picture is called for in the limit ξ→1,where the anti-collinear valence quarks ψand ¯ψcarry almost all of the longitudinal momentum [3].In this case,the residual momentum component (1−ξ)n ·p is small,and the remaining partons are soft-collinear.Each valence quark is described by an anti-collinear jet propagating through the soft-collinear cloud made up of the remaining partons.The two anti-collinear jets communicate through soft-collinear gluon exchange only.The distinct roles played by the valence and remaining partons as ξ→1make it appro-priate to introduce an effective field-theory description for the parton distribution function,in which it is matched onto an operator involving anti-collinear and soft-collinear fields in SCET.The most general,gauge-invariant form the relation (23)can be matched onto in the ξ→1limit readsφns q (ξ,µ)|ξ→1=12(W †¯c ξ¯c )(0)|N (p ) .(24)It is understood that the anti-collinear fields located at the points 0and tn interact only via soft-collinear gluon exchange.Both (24)and the QCD matrix element (23)depend on the single invariant p 2=m 2,so there is no non-trivial hard matching coefficient.The matrix element (24)is precisely the object we encountered in (22).We can use this correspondence along with some simple algebra to findW 1=|C V (Q 2,µ)|2i d (n ·k )¯n ·q J (q 2+n ·k ¯n ·q,µ)φns qn ·kπIm W 1,where the e q are quark electric charges.Inserting the definition of the jet function (21),and recalling that q 2=−Q 2and n ·p ¯n ·q =Q 2/x +power corrections,we obtain the final result for the factorization formulaF ns 2(x,Q 2)= q e 2q |C V (Q 2,µ)|2Q 2 1x dξJ Q 2ξ−x。