1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %% Trim Size: 9.75in x 6.5in
3: %% Text Area: 8in (include Runningheads) x 5in
4: %% ws-ijmpa.tex : 28 July 2003
5: %% Tex file to use with ws-ijmpa.cls written in Latex2E.
6: %% The content, structure, format and layout of this style file is the
7: %% property of World Scientific Publishing Co. Pte. Ltd.
8: %% Copyright 1995, 2002 by World Scientific Publishing Co.
9: %% All rights are reserved.
10: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
11:
12: \documentclass{ws-ijmpa}
13: \usepackage{amsmath}
14:
15:
16: \begin{document}
17:
18: \markboth{Stefan Sint}
19: {The Kaon $B$-parameter in quenched QCD}
20:
21: %%%%%%%%%%%%%%%%%%%%% Publisher's Area please ignore %%%%%%%%%%%%%%%
22: %
23: %\catchline{}{}{}{}{}
24: %
25: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
26:
27: \title{The Kaon $B$-parameter in quenched QCD\footnote{based on work done
28: in collaboration with Petros Dimopoulos, Jochen Heitger, Filippo Palombi,
29: Carlos Pena and Anastassios Vladikas.}}
30:
31:
32: \author{\footnotesize STEFAN SINT (ALPHA-collaboration)}
33:
34: \address{Universidad Aut\'onoma de Madrid, Instituto de F\'{\i}sica
35: Te\'orica,\\
36: Facultad de Ciencias C-XVI, \\
37: E-28049 Cantoblanco, Madrid, Spain\\}
38:
39: \maketitle
40:
41: %\pub{Received (Day Month Year)}{Revised (Day Month Year)}
42:
43: \begin{abstract}
44: I report on a recent determination by the ALPHA
45: collaboration of the kaon $B$-parameter
46: using lattice QCD with Wilson type quarks.
47: An effort is made to control all systematic
48: errors except for the quenched approximation.
49: The preliminary result for the renormalization group invariant
50: parameter is $\hat{B}_K = 0.834(37)$, which translates to
51: $B_K^{\overline{\rm MS}}(2\,{\rm GeV})=0.604(27)$
52: in the ${\overline{\rm MS}}$ scheme with anticommuting $\gamma_5$.
53: %\keywords{Keyword1; keyword2; keyword3.}
54: \end{abstract}
55:
56: \section{Introduction}
57:
58: The kaon $B$-parameter remains an important ingredient
59: in current analyses of CP violation in the Standard
60: Model~\cite{Battaglia:2003in}. It is is defined in QCD with
61: up, down and strange quarks by the matrix element
62: of a four-quark operator between kaon states,
63: \begin{eqnarray}
64: \left\langle \bar{K}^0\vert O^{\Delta S=2}_{\rm (V-A)(V-A)}\vert K^0\right\rangle &=& \frac 83 F_K^2m_K^2 B_K,\nonumber\\
65: O^{\Delta S=2}_{\rm (V-A)(V-A)} &=& \sum_\mu\left[\bar{s}\gamma_\mu(1-\gamma_5)d\right]^2.
66: \end{eqnarray}
67: While this matrix element is well-defined, its relation
68: to the full amplitude in the Standard Model
69: relies on the hypothesis that the charm quark can
70: be treated as a heavy particle.
71: While the quality of this approximation
72: is difficult to assess, it is typically assumed to be valid
73: at the five percent level, which sets
74: the scale for the precision to be attained by
75: a lattice determination of $B_K$.
76:
77: %that a 5 per cent uncertainty due to this approximation
78: %should be assumed, which therefore sets the scale for
79: %the precision to be attained in lattice simulations.
80:
81: \section{$B_K$ and Wilson type quarks}
82:
83: Despite recent progress with other fermion formulations,
84: lattice QCD with Wilson type quarks remains
85: attractive because it is computationally
86: cheap and does not suffer from mixing between flavour and spin
87: degrees of freedom (unlike staggered fermions).
88: On the other hand, all axial symmetries are explicitly
89: broken, which leads to mixing of operators with opposite chirality,
90: and the possible occurence of unphysical fermion
91: zero modes in the quenched approximation.
92: In order to avoid the latter problem in typical
93: quenched simulations, the masses of pseudoscalar
94: mesons are typically heavier than the physical kaon mass.
95: Concerning the calculation of $B_K$, the mixing problem
96: is particularly annoying. Decomposing the
97: operator $O_{\rm (V-A)(V-A)}=O_{\rm VV+AA}-O_{\rm VA+AV}$
98: the relevant piece for $K^0$-$\bar{K}^0$ mixing is the parity even
99: component, which renormalizes as follows~\cite{Bernard:1987pr}:
100: \begin{eqnarray}
101: \left[O_{\rm VV+AA}\right]_{\rm R} &=&
102: Z_{\rm VV+AA}\left[O_{\rm VV+AA}+\sum_{i=1}^4 z_i\ O_i^{d=6} \right].
103: \end{eqnarray}
104: While the mixing problem can be solved by imposing axial continuum
105: Ward identities~\cite{Aoki:1999gw},
106: it represents a major obstacle for precision results.
107: Note that the parity-odd operator component $O_{\rm VA+AV}$
108: does indeed renormalize multiplicatively~\cite{Bernard:1987pr}.
109: This can be exploited for the computation of $B_K$ as will be explained
110: shortly.
111:
112: \section{QCD with chirally rotated mass terms}
113:
114: Axial and vector symmetries can be distinguished according
115: to whether or not the quark mass term is left invariant.
116: Hence, a non-standard form of the quark mass term
117: also modifies the form of the symmetry transformations.
118: Let us consider the continuum theory for a light quark doublet $\psi$
119: and the $s$ quark including a chirally twisted mass term,
120: \begin{equation}
121: {\cal L}_f=\overline{\psi}\left(D\kern-6pt\slash
122: +m+i\mu_q\gamma_5\tau^3\right)\psi
123: + \bar{s}\left(D\kern-6pt\slash +m_{\rm s}\right)s.
124: \end{equation}
125: After a chiral rotation of the doublet fields
126: \begin{equation}
127: \psi' =\exp\left(i \alpha\gamma_5\frac{\tau^3}{2}\right)\psi,\qquad
128: \overline{\psi}'
129: =\overline{\psi}\exp\left(i \alpha\gamma_5\frac{\tau^3}{2}\right),
130: \end{equation}
131: and with $\alpha$ chosen such that $\tan\alpha = \mu_{\rm q}/m$,
132: the Lagrangian reads
133: \begin{equation}
134: {\cal L}'_f=\overline{\psi}'\left(D\kern-6pt\slash +m'\right)\psi'
135: + \bar{s}\left(D\kern-6pt\slash +m_{\rm s}\right)s,\qquad
136: m'= \sqrt{m^2+\mu_q^2}.
137: \end{equation}
138: The field rotation re-defines the symmetries and maps composite fields, e.g.
139: \begin{equation}
140: O_{\rm VV+AA}' = \cos(\alpha)O_{\rm VV+AA} - i\sin(\alpha)
141: O_{\rm VA+AV}
142: = -i O_{\rm VA+AV}\quad (\alpha=\pi/2).
143: \end{equation}
144: In particular, the operators are mapped to each other at
145: ``maximal twist'' $\alpha=\pi/2 \Leftrightarrow m=0$,
146: where the quark mass is determined entirely by the
147: chirally twisted mass parameter $\mu_{\rm q}$.
148: Hence, by using Wilson quarks with a maximally twisted
149: light quark doublet and a standard $s$ quark, the complicated operator
150: mixing problem can be by-passed~\cite{Frezzotti:2000nk}.
151: An additional benefit consists in the elimination of
152: unphysical zero modes by the twisted mass term.
153: This allows for numerical simulations to get close
154: to the physical situation: light, mass degenerate $u,d$ quarks
155: and a heavier $s$ quark. However,
156: most results in lattice QCD are
157: currently obtained in the limit where the kaon is made
158: out of mass-degenerate $d$ and $s$ quarks. In this way
159: a quenched artefact is avoided~\cite{Sharpe:1994dc}, but
160: it implies that the zero mode problem is back for the
161: standard Wilson $s$ quark.
162: In order to cope with mass-degenerate quarks
163: we also consider a different set-up, where
164: the r\^oles of $s$ and $u$ quarks are interchanged.
165: The chirally twisted doublet is now $\psi=(s,d)$
166: and after the chiral rotation one finds
167: \begin{equation}
168: O_{\rm VV+AA}' = \cos(2\alpha)O_{\rm VV+AA}
169: - i\sin(2\alpha) O_{\rm VA+AV}
170: = -i O_{\rm VA+AV}\quad (\alpha=\pi/4).
171: \end{equation}
172: Again the operators are mapped to each other provided
173: $\alpha=\pi/4 \Leftrightarrow m=\mu_{\rm q}$. For
174: obvious reasons we will refer to the two set-ups
175: as $\pi/2$ and $\pi/4$ scenarios, respectively.
176: The latter has the advantage that both
177: $d$ and $s$ quark masses can be decreased simultaneously
178: without encountering unphysical zero modes.
179: Setting $\alpha$ to $\pi/2$ or $\pi/4$ requires some
180: parameter tuning using known results for
181: finite renormalization constants. In addition one needs the
182: multiplicative operator renormalization constant.
183: The renormalization problem has been solved
184: non-perturbatively in~\cite{Guagnelli:2005zc}, using a finite volume
185: scheme based on the Schr\"odinger functional
186: and recursive finite size techniques~\cite{Jansen:1995ck}.
187:
188: \section{Numerical simulations}
189: The numerical simulations have been performed using the O($a$) improved Wilson
190: quark action and the Wilson's plaquette action for the gauge fields.
191: 4-5 $\beta$-values have been chosen in the interval $[6.0-6.45]$,
192: corresponding to lattice spacings $a=0.05-0.1\,{\rm fm}$, if the
193: scale is set by $r_0=0.5\,{\rm fm}$.
194: The lattice volumes range form $16^3\times 48$ to $32^2\times 72$.
195: Quark masses are tuned to achieve $\alpha=\pi/2$ or $\alpha=\pi/4$
196: and pseudoscalar masses around or above $m_K$.
197: The analysis of excited states determined safe plateaux regions
198: where $B_K$ could be extracted from suitable ratios
199: of correlation functions.
200: Finite volume effects were checked at the coarsest lattice spacing
201: ($\beta=6.0$), and found to be below the statistical errors.
202: In the $\pi/2$ scenario chiral extrapolations were performed linearly
203: to the physical kaon mass. In the $\pi/4$ scenario
204: the kaon mass could be reached by {\em interpolation}
205: except at $\beta=6.45$ where finite volume effects at the
206: physical kaon mass would have been non-negligible.
207: The resulting values for the $B_K$ are given in figure~1 as a function
208: of $a/r_0$. Discarding the values at the coarsest lattice the
209: data seem to scale very well. A continuum extrapolation of $\pi/4$
210: data linear in $a$ seems adequate.
211:
212: %\psdraft
213: \begin{figure}
214: \centerline{
215: \psfig{file=./BKrgi.eps,
216: width=14cm,bbllx=18,bblly=144,bburx=592,bbury=490}}
217: \vspace*{-1.5cm}
218: \caption{$B_K$ data for both $\pi/4$ and $\pi/2$ scenarios. The $\pi/2$
219: data is slightly shifted to the right to enhance readibility. Also shown
220: is the linear continuum extrapolation of $\pi/4$ data excluding the data on the
221: coarsest lattice.}
222: \end{figure}
223:
224: \section{Conclusions}
225:
226: We have performed a benchmark calculation of $B_K$ where,
227: apart from the quenched approximation, all systematic
228: errors are under control. These include cutoff effects,
229: finite volume effects and the contamination by excited states.
230: A non-perturbative renormalization procedure has been employed,
231: and the continuum limit has been taken.
232: The preliminary result
233: \begin{equation}
234: \hat{B}_K = 0.834(37) \qquad \Leftrightarrow \qquad
235: B_K^{\overline{\rm MS}}(2\,{\rm GeV})=0.604(27),
236: \end{equation}
237: has the desired precision. Within errors it is
238: compatible with results obtained by other groups using
239: different lattice regularisations
240: (see \cite{Dawson} for a recent review and further references).
241: Hence, further progress will require
242: the inclusion of dynamical quark flavours. The recent progress
243: with simulation algorithms for Wilson-type quarks~\cite{Luscher:2005mv}
244: will soon allow to compute $B_K$ with a similar precision
245: including the effect of light dynamical up and down quarks.
246:
247: \section*{Acknowledgments}
248:
249:
250: I would like to thank the organisers for the invitation
251: to this nice conference, my collaborators
252: P.~Dimopoulos, C.~Pena,~F. Palombi,
253: A.~Vladikas, J.~Heitger for a pleasant collaboration,
254: and the computer centre at DESY Zeuthen for providing
255: the necessary CPU time on the APEmille machines.
256:
257: %\section*{References}
258:
259: \begin{thebibliography}{99}
260:
261: %\cite{Battaglia:2003in}
262: \bibitem{Battaglia:2003in}
263: M.~Battaglia {\it et al.},
264: %``The CKM matrix and the unitarity triangle,''
265: arXiv:hep-ph/0304132.
266: %%CITATION = HEP-PH 0304132;%%
267:
268: %\cite{Bernard:1987pr}
269: \bibitem{Bernard:1987pr}
270: C.~W.~Bernard, T.~Draper, G.~Hockney and A.~Soni,
271: %``Recent Developments In Weak Matrix Element Calculations,''
272: Nucl.\ Phys.\ Proc.\ Suppl.\ {\bf 4} (1988) 483.
273: %%CITATION = NUPHZ,4,483;%%
274:
275: %\cite{Aoki:1999gw}
276: \bibitem{Aoki:1999gw}
277: S.~Aoki {\it et al.} [JLQCD Collaboration],
278: %``The kaon B-parameter with the Wilson quark action using chiral Ward
279: %identities,''
280: Phys.\ Rev.\ D {\bf 60} (1999) 034511
281: [arXiv:hep-lat/9901018].
282: %%CITATION = HEP-LAT 9901018;%%
283:
284: %\cite{Frezzotti:2000nk}
285: \bibitem{Frezzotti:2000nk}
286: R.~Frezzotti, P.~A.~Grassi, S.~Sint and P.~Weisz [Alpha collaboration],
287: % ``Lattice QCD with a chirally twisted mass term,''
288: JHEP {\bf 0108} (2001) 058 [arXiv:hep-lat/0101001].
289: %%CITATION = HEP-LAT 0101001;%%
290:
291: %\cite{Sharpe:1994dc}
292: \bibitem{Sharpe:1994dc}
293: S.~R.~Sharpe,
294: %``Phenomenology from the lattice,''
295: arXiv:hep-ph/9412243.
296: %%CITATION = HEP-PH 9412243;%%
297:
298:
299: %\cite{Guagnelli:2005zc}
300: \bibitem{Guagnelli:2005zc}
301: M.~Guagnelli, J.~Heitger, C.~Pena, S.~Sint and A.~Vladikas [ALPHA
302: Collaboration],
303: % ``Non-perturbative renormalization of left-left four-fermion operators in
304: % quenched lattice QCD,''
305: arXiv:hep-lat/0505002.
306: %%CITATION = HEP-LAT 0505002;%%\end{thebibliography}
307:
308: %\cite{Jansen:1995ck}
309: \bibitem{Jansen:1995ck}
310: K.~Jansen {\it et al.},
311: % ``Non-perturbative renormalization of lattice QCD at all scales,''
312: Phys.\ Lett.\ B {\bf 372} (1996) 275
313: [arXiv:hep-lat/9512009].
314: %%CITATION = HEP-LAT 9512009;%%
315:
316:
317: \bibitem{Dawson}
318: C. Dawson, plenary talk at the XXIIIrd International Symposium on
319: Lattice Field Theory, 25-30 July 2005, Trinity College, Dublin, Ireland
320:
321: %\cite{Luscher:2005mv}
322: \bibitem{Luscher:2005mv}
323: M.~L\"uscher,
324: %``Lattice QCD with light Wilson quarks,''
325: arXiv:hep-lat/0509152.
326: %%CITATION = HEP-LAT 0509152;%%
327:
328: \end{thebibliography}
329:
330: \end{document}
331:
332:
333: