1: % Chapter 5
2:
3: \section{Summary}\label{conclusion1}
4:
5: A great effort has been devoted in the last few years in order to
6: develop a deeper qualitative and quantitative understanding of systems
7: described by interacting quantum fields out of equilibrium. There is a
8: class of physical problems that requires the consistent treatment of
9: time dependent mean--fields in interaction with their own quantum or
10: thermal fluctuations. I may mention, among others, the problem of
11: reheating of the universe after the inflationary era of exponential
12: growth and cooling, and the time evolution of the scalar order
13: parameter through the chiral phase transition, soon to be probed in
14: the heavy--ion experiments at BNL--RHIC and CERN--LHC. In these
15: situations, a detailed description of the time--dependent dynamics is
16: necessary to calculate the non--equilibrium properties of the
17: system. Indeed, the development of practical general techniques and
18: the advent of faster and cheaper computers have made possible the
19: discovery of novel and unexpected phenomena, ranging from dissipative
20: processes via particle production to novel aspects of symmetry
21: breaking
22: \cite{Cooper:1997ii,Boyanovsky:1995me,Boyanovsky:1998ba,Boyanovsky:1998zg}.
23:
24: From the technical point of view, it should be pointed out, first of
25: all, that a perturbative treatment of this dynamical problem is
26: meaningful only when the early time evolution is considered. The
27: presence of parametric resonant bands or spinodal instabilities (in
28: the case, respectively, of unbroken or spontaneously broken
29: symmetries) rapidly turns the dynamics completely non--linear and
30: non--perturbative. Thus, the asymptotic evolution at late time can be
31: consistently studied only if approximate {\em non--perturbative}
32: methods are applied to the problem \cite{Boyanovsky:1995me}.
33:
34: Quite recently one of these schemes, namely the large $N$ expansion at
35: leading order (LN) \cite{largen_exp,Cooper:1997ii}, has been used in
36: order to clarify some dynamical aspects of the $\phi^4$ theory in $3$
37: spatial dimensions, reaching the conclusion that the non--perturbative
38: and non--linear evolution of the system might eventually produce the
39: onset of a non--equilibrium Bose--Einstein condensation (BEC) of the
40: long--wavelength Goldstone bosons usually present in the broken
41: symmetry phase
42: \cite{Boyanovsky:1998ba,Boyanovsky:1998yp,Boyanovsky:1998zg}. Another
43: very interesting result in \cite{Boyanovsky:1998yp} concerns the
44: dynamical Maxwell construction, which reproduces the flat region of
45: the effective potential in case of broken symmetry as asymptotic fixed
46: points of the background evolution.
47:
48: In section \ref{numerical} (see also ref. \cite{Destri:1999hd}) I have
49: addressed the question of whether a standard BEC could actually take
50: place as time goes on, by putting the system in a finite volume (a
51: periodic box of size $L$) and carefully studying the volume dependence
52: of out--of--equilibrium features in the broken symmetry phase. We
53: summarize here the main results contained in section \ref{numerical} as
54: well as in \cite{Destri:1999hd}. The numerical solution shows the
55: presence of a time scale $\tau_L$, proportional to the linear size $L$
56: of the system, at which finite volume effects start to manifest, with
57: the remarkable consequence that the zero-mode quantum fluctuations
58: cannot grow macroscopically large if they start with microscopic
59: initial conditions. In fact, the size of low--lying widths at time
60: $\tau_L$ is of order $L$, to be compared to order $L^{3/2}$ for the
61: case of standard BEC. In other words we confirmed that the linear
62: growth of the zero mode width, as found also by the authors of
63: \cite{Boyanovsky:1998ba,Boyanovsky:1998yp,Boyanovsky:1998zg}, really
64: signals the onset of a novel form of dynamical BEC, quite different
65: from the standard one described by equilibrium finite--temperature
66: field theory. This interpretation is reinforced by the characteristics
67: of the long--wavelength fluctuations' spectrum.
68:
69: To go beyond the gaussian approximation and study the interaction
70: between longitudinal and transverse fluctuations, I have defined in
71: section \ref{impHF} an extended time dependent Hartree-Fock (tdHF)
72: approximation for the $\phi^4$ QFT, which includes some non-gaussian
73: features of the full theory (for an introduction to the standard
74: tdHF, see ref. \cite{Kerman:1976yn}). I have presented a rather
75: detailed study of the dynamical evolution out of equilibrium, in
76: finite volume (a cubic box of size $L$ in $3$D), as well as in
77: infinite volume. For comparison, I have also analyzed some static
78: characteristics of the theory both in unbroken and broken symmetry
79: phases.
80:
81: By means of a proper substitution of the bare coupling constant with
82: the renormalized coupling constant (fully justified by diagrammatic
83: consideration), I have been able to obtain equations of motion
84: completely independent of the ultraviolet cut-off (apart from a slight
85: dependence on inverse powers, that is, however, ineluctable because of
86: the Landau pole). I have described in detail the shape of the ground
87: state, showing how a broken symmetry scenario can be recovered from
88: the quantum mechanical model, when the volume diverges.
89:
90: Moreover, I have shown that, within this slightly enlarged tdHF
91: approach that allows for non--gaussian wavefunctions, one might
92: recover the usual gaussian HF approximation in a more controlled way.
93: In fact, studying the late time dynamics, I have confirmed the
94: presence of a time scale $\tau_L$, proportional to the linear size $L$
95: of the box, at which the evolution ceases to be similar to the
96: infinite volume one. At the same time, the low--lying modes amplitudes
97: have grown to order $L$. The same phenomenon has been observed in the
98: $O(N)$ model \cite{Destri:1999hd}. Looking at this result in the framework of
99: our extended tdHF approximation, one realizes that the growth of
100: long--wavelength fluctuations to order $L$ in fact undermines the
101: self--consistency of the gaussian HF itself. In fact, in our tdHF
102: approach the initial gaussian wavefunctions are allowed to evolve into
103: non--gaussian forms, but they simply do not do it in a macroscopic
104: way, within a further harmonic approximation for the evolution, so
105: that in the infinite--volume limit they are indistinguishable from
106: gaussians at all times. But when $M^2$ is on average not or order
107: $L^0$, but much less, as it happens for suitable initial conditions,
108: infrared modes of order $L$ will be dominated by the quartic term in
109: our Schroedinger equations (\ref{Schroedinger}), showing a possible
110: internal inconsistency of the gaussians approximation.
111:
112: An other manifestation of the weakness of the HF scheme is the curious
113: ``stopping at the spinodal line'' of the width of the gaussian quantum
114: fluctuations, when the initial configuration does not break the
115: symmetry. This does not happen in the large $N$ approach because of
116: different coupling of transverse modes (the only ones that survive in
117: the $N\to\infty$ limit) with respect to the longitudinal modes of the
118: $N=1$ case in the HF approach.
119:
120: I have also described the non--trivial phenomenology of the
121: infinite--volume late--time evolution in the gaussian approximation,
122: showing how the dynamical Maxwell construction differs from the
123: $N=\infty$ case. In fact, I have observed the presence of an unstable
124: interval, contained in the static flat region which is forbidden as
125: attractor of the asymptotic evolution. This region corresponds, more
126: or less, to the spinodal region of the classical potential, with the
127: obvious exception of the origin. In particular, I have found that the
128: energy flux between the classical degree of freedom and the bath of
129: quantum fluctuations is quite complex and not monotonous. In other
130: words, since I start from initial conditions where the fluctuation
131: energy is not minimal, there are special situations where enough
132: energy is transferred from the bath to the condensate, pushing it
133: beyond the top of the potential hill.
134:
135: As far as the non linear $\sigma$ model in $(1+1)$D is concerned, I
136: have shown explicitly in chapter \ref{smodel} that the large $N$ limit
137: and the large coupling limit, which turns the linear model in the non
138: linear one, commute. I have also derived the evolution equations for
139: the non linear model at the leading order in the $1/N$ expansion, in
140: the case of a field condensate different from zero. I implemented the
141: constraint by the use of a Lagrange multiplier, which I denoted $m^2$,
142: since it enters the dynamics as a squared mass. I have shown that the
143: usual renormalization procedure, which makes the bare coupling
144: constant depend on the UV cutoff, is sufficient to get properly
145: renormalized, that is UV finite, evolution equations. Moreover, I
146: characterized the ground state of the model, giving an interpretation
147: of the dynamical generation of mass (the so--called dimensional
148: transmutation) in terms of a compromise between energetic requirements
149: and the constraint. I then solved numerically the evolution equations
150: and analyzed the history of the condensate and the Lagrange multiplier
151: as well as the number of particles created during the relaxation of
152: the condensate (the quantum fluctuations). Remarkably, I do not find
153: any period of exponential growth for the fluctuations. Actually, no
154: spinodal instabilities were to be expected, since the symmetry is
155: always unbroken in $1+1$ dimensions. But there occurs also no parametric
156: resonance, as takes place instead in the unbroken symmetry scenario of
157: the large$-N$ $\phi^4$ model in $3+1$ dimensions.
158:
159: The estimated dependence of the asymptotic value of the Lagrange
160: multiplier, $m^2(\infty)$, on the initial condensate $\rho_0$, turns
161: out to be very well approximated by an exponential, which is the exact
162: dependence of $m^2(0)$ [at infinite UV cutoff, see
163: eqs. (\ref{alfadiro}) and (\ref{m0dia})]; remarkably however, the
164: prefactor in the exponent is changed [see eq. (\ref{minfrho})]. As far
165: as the emission of particles is concerned, we considered three
166: different reference states: the initial state, the adiabatic vacuum
167: state and the equilibrium vacuum state, that is the true ground state
168: of the theory. The numerical results suggest a weak non linear
169: resonance, yielding a relaxation of the condensate via particle
170: production driven by power laws with non universal anomalous
171: exponents, a result similar to what found in \cite{Boyanovsky:1998zg}
172: for the asymptotic dynamics of $\phi^4$ in $3+1$ dimensions. Finally,
173: since I allow the condensate to have a number $n$ of components larger
174: than $1$, I was able to study the evolution of configurations with
175: non--zero angular momentum $\ell$ in the internal space of the field
176: [see eq. (\ref{eqm1})]. In this case I find numerical evidence for an
177: adiabatic spectrum broader than in the case $\ell=0$ (see figure
178: \ref{lneq0_3}), suggesting a stronger coupling with hard modes.
179:
180:
181:
182: \section{Open issues}
183:
184: I close this section and this work giving a list of the (in my
185: opinion) most relevant issues still open, whose solution may result in a
186: better understanding of both the formal and phenomenological aspects
187: of QFT out of equilibrium:
188:
189: \begin{enumerate}
190: %
191: \item The study of the Poincar\'e cycles in the effective hamiltonian
192: dynamics and their dependence on the parameters
193: \cite{Cooper:1997ii,Destri:1999hd} may allow for a deeper understanding of the
194: dynamics in finite volume.
195: %
196: \item The inclusion of $1/N$ corrections, which contain collisional
197: contributions, is needed to better understand the issue of
198: thermalization; they, in fact, may lead to a thermal asymptotic
199: distribution and make the damping more efficient
200: \cite{Bettencourt:1998xb,Bettencourt:1998nf}. With this regard, it
201: should be stressed that finite truncations of the large--$N$ expansion
202: introduce errors which grow with time and make the approximation
203: unreliable on a time scale proportional to as $N^{1/2}$ \cite{Ryzhov:2000fy}.
204: %
205: \item Further study, both analytical and numerical, is needed in
206: the generalized tdHF approach of section \ref{impHF}, to better
207: understand the dynamical evolution of quantum fluctuations in the
208: broken symmetry phase coupled to the condensate. An interesting
209: direction is the investigation of the case of finite $N$, in order to
210: interpolate smoothly the results for $N=1$ to those of the $1/N$ (at
211: leading order) approach. It should be noted, in fact, that the theory
212: with a single scalar field contains only the longitudinal mode (by
213: definition), while only the transverse modes are relevant in the large
214: $N$ limit. Hence a better understanding of the coupling between
215: longitudinal and transverse modes is necessary. In this direction,
216: another relevant point is whether the Goldstone theorem is respected
217: in the HF approximation
218: \cite{Lenaghan:1999si}. It would be interesting also to study the dynamical
219: realization of the Goldstone paradigm, namely the asymptotic vanishing
220: of the effective mass in the broken symmetry phases, in different
221: models; this issue needs further study in the $2D$ case \cite{Cooper:1997ii},
222: where it is known that the Goldstone theorem is not valid.
223: %
224: \item The issue of thermalization may be addressed also by deriving a
225: Quantum Boltzmann (kinetic) equation. The issue is relevant
226: for the reheating problem and for the dynamics of heavy ion
227: collisions. It would be useful a systematic quantal generalization of
228: classical transport theory, in the presence of strong inhomogeneous
229: mean fields, where quantum self--energy (off--shell virtual processes)
230: are included on the same footing as collisional or real particle
231: production processes in the plasma. To this end, the CTP generating
232: functional may be used in order to derive an effective action
233: principle which leads to a hierarchy of Schwinger--Dyson equations
234: \cite{Witte:1994an}. Some preliminary results \cite{Boyanovsky:1998pg}
235: shows an anomalous relaxation of hard quasiparticles. Considering
236: higher orders will include collisional processes which will give
237: exponential decay in time. The competition between collisional and
238: Landau damping time scales will then depend on the particular model
239: under study.
240:
241: In any case, the approach to thermodynamic equilibrium is far from
242: being a trivial subject. In fact, even though the time evolution of
243: correlation functions can be described by an exact evolution
244: (functional) equation for the corresponding generating functionals
245: \cite{Wetterich:1997ap} (which shows many dynamical features of
246: quantum mechanics
247: \cite{Wetterich:1997zt}) and the classical field equations admit
248: thermodynamic equilibrium as a fixed point, infinitely many conserved
249: correlation functions prevent the system from approaching the fixed
250: point, when it starts in its neighborhood. Thus, it seems that
251: equilibrium can therefore be reached at most for suitably averaged
252: quantities or for subsystems. A similar analysis in QFT leads to the
253: same conclusion that a uniform approach to thermal equilibrium is
254: prevented by the existence of infinitely many other fixed points which
255: correspond to incoherent mixtures of eigenstates of the quantum
256: hamiltonian \cite{Wetterich:1997rp}.
257:
258: Still on this subject, it is worth noticing that an important
259: understanding both on thermalization and on the quantal generalization
260: of kinetic equations, has been reached recently by the authors of
261: ref.s \cite{Berges:2000ur} and \cite{Aarts:2001qa}. Using a three-loop
262: approximation to th 2PI effective action and solving numerically the
263: evolution equations for the two point function in $1+1$ dimensions, it
264: is possible to show that all correlation functions approach the
265: thermal distribution asymptotically, proving that higher loops contain
266: indeed the collisional contributions.
267:
268: %
269: \item The study of Disoriented Chiral Condensates has been started
270: modelling the pion phenomenology with the $O(4)$ linear $\sigma$
271: model. It would be interesting to look for qualitative and
272: quantitative differences in the predictions, when they are extracted
273: from the non-linear model.
274: %
275: \item To better understand out of equilibrium phenomena in many areas
276: of physics, ranging from condensed matter and statistical mechanics
277: systems to cosmology and particle physics, it is necessary to follow
278: the evolution of non-homogeneous condensates in several models. The
279: derivation of the evolution equations for a spherically symmetric
280: condensate, coupled to its quantum fluctuations, I gave in section
281: \ref{dis_homo}, is a promising first step in this direction and deserves a
282: deeper analytical and numerical analysis, in view also of its
283: phenomenological implications. This approach may be married with the
284: the technique developed by Aarts and Smit \cite{Aarts:1998zu},
285: \cite{Aarts:1998gi}, \cite{Aarts:1998td} and \cite{Aarts:1999zn} in
286: order to include the quantum fluctuations of both fermionic and
287: bosonic fields, with inhomogeneous condensates.
288: %
289: \item I commented a lot in chapter \ref{gauge} about the results and
290: the open problems in non equilibrium gauge theories: they are relevant
291: for our knowledge of phase transitions in Standard Model and
292: baryogenesis
293: \cite{Riotto:1998bt,Buchmuller:2000wq}. Domain walls (non homogeneous
294: field configurations) for the non abelian gauge fields of the
295: electroweak theory (usually known with the name of {\em sphalerons})
296: are thought to be responsible of the matter-antimatter asymmetry we
297: see in our universe.
298: %{\em maybe transfer here chapter 4 ??}
299: Thus, in addition to the issue raised in the previous point, we need
300: to find a dynamical scheme suitable to gauge theories. To the best of
301: my knowledge, a non perturbative treatment in the gauge coupling is
302: still lacking. One may try to define a mean field approximation by
303: means of a time--dependent variational principle. It is still not
304: clear, however, whether one should introduce the gauge conditions as
305: constraints in the variational principle or one should enforce these
306: conditions in the choice of the trial space \cite{Benarous:1997kp}
307: %{\em add problems}
308:
309: Finally, detailed computation of transport coefficients, damping rates
310: and energy loss characteristics of the quark gluon plasma from the
311: fundamental microtheory, QCD \cite{QFTooE3} is also needed, in view of
312: the forthcoming experimental results from BNL and CERN.
313: \end{enumerate}
314:
315:
316:
317:
318:
319:
320:
321:
322: