INTRODUCTION 1 A possible quantum probability in rease of the ylindri al gravitational (cid:28)eld D. Bar 7 0 Abstra t 0 2 r As known, the ylindri al gravitational (cid:28)eld (wave) have been anoni ally quantized a M and its wave fun tion, as the quantum one, interpreted in probability terms. We show 1 in this work, using quantum Zeno methods, that this probability may be substantially 2 2 in reased and even approa h unity. For that we (cid:28)rst show, in detailed manner, that the v 7 ylindri al gravitational (cid:28)eld may be dis ussed in the ommutation number representa- 8 0 tion. We also dis uss this (cid:28)eld in the transverse-tra eless (TT) gauge and al ulate the 3 0 related trapped surfa e. 7 0 / c q - r Key words: Cylindri al gravitational waves, Quantization, Zeno e(cid:27)e t g : v Pa s numbers: 03.70.+k, 04.30.-w, 03.65.Xp i X r a 1 INTRODUCTION 1 The problem of quantizing the general gravitational (cid:28)eld (GF) (gravitational wave (GW) [1℄) have, theoreti ally, been ta kled by di(cid:27)erent methods and persons beginning from the earlier works of Rosenfeld [2℄, Bergmann [3℄ and S hwinger [4℄ to the anoni al methods of Dewitt [5℄, Adm [6℄ and Dira [7℄. Among the supposed quantum hara teristi s of the GF (GW) is the probability interpretation [8, 9℄ whi h is extended from the quantum regime so that one interprets [10℄ the GF (GW) fun tion as a probability amplitude [8, 9℄. This problem of quantizing and interpreting, in probabilityterms, the GF entails in turn a se ond interesting problem of how and in whi h way to in rease the probabilty of some spe i(cid:28) GF. 1 Sin e the problemof quantizationis found in the literature to refer to gravitational(cid:28)eld [4℄ and alsoto gravitational wave[10℄ we use here these two on epts side by side or inter hangably on equal basis. INTRODUCTION 2 Taking the ylindri al GF as an example and assuming that one starts from a hypersurfa e with a ylindri al geometry ane may ask how to in rease its probability so that this same ylindri al geometry may persist upon this hypersurfa e. This problem is also found, under di(cid:27)erent terms and terminology, in other dis iplines of generalrelativity. Thus, a entralissueofquantumgravity[11℄istheproblemofsubstantially in reasing the probability of some quantum foam [12℄, with a typi al order of magnitude of G¯h the Plan k-Wheeler length ( c3 ) [13, 14℄, so that it may be realized in full grown human s ale with a unity probability. An example of the latter, whi h is intensely dis ussed in the literature [12, 15℄, is that of the theoreti al in rease of some foam-like probability tissue in the form of wormhole to the realized human-s ale (of unity probability) so that it may be used for spa etime and, espe ially, time travelling [12, 15℄. Asknown, thementioned(cid:28)rstproblemofquantizingtheGF(GW)(cid:28)ndsitssolutionwhen one restri ts the dis ussion to limited dynami al regions of geometrodynami s su h as the minisuperspa e and miniphasespa e dis ussed by Dewitt [5℄, Misner [16℄ and, espe ially, by Ku har [10℄ whi h applied it for quantizing the ylindri al GW. Ku har did it by extending the extrinsi time idea, whi h was (cid:28)rst introdu ed [6, 17, 18℄ for the linearized theory [13, 14℄ of general relativity, to the nonlinear theory [19℄ and, espe ially, to the ylindri al spa etime. This extrinsi time variable, whi h is of the Tomonaga-S hwinger many-(cid:28)ngered time kind [20, 21℄, is anoni ally onjugate to momentum and not to energy as is the intrinsi time whi h serves more as a label to distinguish between spa elike hypersurfa es [13℄ in a one- parameter family of them. Thus, using the extrinsi time variable one may obtain [10℄, as known from [6℄, a formal- ism whi h is identi al to that used for dis ussing the parametrized [10℄ ylindri al massless s alar (cid:28)eld in a (cid:29)at Minkowskian ba kground. This suggests, as noted in [10℄, that all the results obtained from the later ase may, theoreti ally, be applied also for the ylindri al GF in a urved spa etime. These appli ations in lude also the anoni al quantization of the ylindri al GF (GW) and the derivation [10℄ of a fun tional time-independent S hroedinger- INTRODUCTION 3 type equation for it. Moreover, Ku har has su eeded [10℄ to apply for the ylindri al GF (GW) not only quantum ideas su h as the inner produ t of states [8, 9℄ or path independen e of the evolution of them but also their probability interpretation [8, 9℄. In this work we use this su essful quantization of the ylindri al GW as a basis for dis- ussing the mentioned se ond problem of (cid:28)nding the onditions under whi h the mentioned probabilityin reases andevenapproa hes unity. In thequantum regime,where states hange with both time and spa e [8, 9℄, there exists the known Zeno e(cid:27)e t [22, 23, 24, 25℄ whi h auses these hanging states to be ome onstant and (cid:28)xed and, therefore, to ause their probability, as in lassi al physi s, to be ome unity. This e(cid:27)e t, whi h were experimentally validated [26, 27℄, has three di(cid:27)erent versions: (1)Repeatinga largenumber of times, ina (cid:28)nitetotaltime, the same experiment of he king the present state of some quantum system that has been prepared in an initial spe i(cid:28) state so that in the limit of a very large number of repetitions in the same total time the initial state is preserved in time [22, 26℄. (2) Performing a large set of di(cid:27)erent experiments ea h redu ing the system to some state (froma large number of di(cid:27)erent ones) thereby obtaininga largenumber of di(cid:27)erent possible paths of states (Feynman paths [28, 29℄) so that in the limit of doing these experiments in a (cid:16)dense(cid:17) manner one may (cid:16)realize(cid:17) [23, 24℄ any spe i(cid:28) path of states in the sense that the probability to pro eed along all its onstituent states tends to unity. (3) Simultaneously performing the same experiment in a large number of non-overlapping spatial subregions all in luded in a (cid:28)nite total region so that when these subregions in(cid:28)nites- imally shrink, keeping the total region (cid:28)xed, one avoid any spatial shifting of the state [25℄. This may be explained by the example of trying to lo ate a very small parti le-like obje t in the (cid:28)nite total spatial region whi h be ome easier when this region is divided into several equal parts ea h o upying the same small obje t and the sear hing for it is done in ea h of these smaller regions. It is obvious that the smaller be ome these regions in the (cid:28)nite total region the probability to lo ate the small obje t in ea h of them grows so that in the limit INTRODUCTION 4 in whi h they in(cid:28)nitesimaly shrink this probability tends to unity. The (cid:28)rst two ases (1)-(2) are, respe tively, known [23, 24℄ as the stati and dynami Zeno e(cid:27)e ts while the third (3) is the the spa e Zeno e(cid:27)e t [25℄. We note that whereas the quantum states are related to the ordinary intrinsi time the gravitational ylindri al states are hara terized by [10℄ an extrinsi time variable whi h is related (and a tually borrows [10℄ its name) to the extrinsi urvature. That is, the grav- itational state, represented by the probability that the related hypersurfa e has ylindri al geometry, hanges in spa etime with respe t to extrinsi time. Thus, sin e this extrinsi time, like the ordinary spatial variable, is [10℄ anoni ally onjugate to momentum (whi h is onne ted to extrinsi urvature) the orresponding gravitational Zeno e(cid:27)e t should also have spatial hara teristi s as in the mentioned spa e Zeno e(cid:27)e t. We dire tly show in this work, using spa e Zeno terms [25℄, that one may avoid any spa e shifting of the ylindri al GF (GW) thereby (cid:28)xing its ylindri al geometry and ausing its probability to approa h unity. In the following we pre ede this Zeno demonstration with a dis ussion wi h shows that the ylindri al GF may be represented as a large number of onstituent parts in some (cid:28)nite region of spa e so that it may be dis ussed in spatial Zeno terms [25℄. The appropriate representation whi h enables one to do so is the o upation number one [8, 9℄. Thus, we dis uss here, in detail, this representation [8, 9℄ in relation to the ylindri al GF. We note in this respe t that although this representation in the ontext of ylindri al GF is mentioned in the literature [10, 30℄ no expli it, as far as we know, and detailed expressions of it exists so far As known from quantum me hani s in the o upation number representation [8, 9, 31℄, one may, theoreti ally, prepare any quantum state by merely applying the relevant reation and destru tion operators [8, 9℄ any required number of times upon some initial basis state. One also knows from the anoni al formalism [6, 13, 14℄ of general relativity that it is possible to prepare the geometry of some spa etime hypersurfa e by ontrolling the form of the lapse and shift fun tions [6, 13, 14℄. That is, any spe i(cid:28) evolution of spa etime should INTRODUCTION 5 be pre eded by determining beforehand these fun tions so that one an be sure (with a unity probability) that the related spa etime is developed along the spe i(cid:28)ed route. This operation, hara terized by the determination of the lapse and shift fun tions, should be related, when dis ussing the quantum properties of the GF, to the orresponding operation of the mentioned quantum reation and destru tion operators upon some initial basis state. We derive here in detail the appropriate expressions whi h a ordingly relate the ylindri al lapse and shift fun tions to the reation and destru tion operators. When the mentioned probability of the ylindri al GW approa h unity (also in onse- quen e of the spatial Zeno e(cid:27)e t) the related GW produ e ertain e(cid:27)e ts upon the neigh- bouring spa etime through whi h it pro eeds su h as implanting its ylindri al geometry upon it and giving rise to some trapped surfa e [32, 33, 34, 35℄. We follow the develope- ment of this ylindri al GW, on e its probabilisti han es were greatly in reased, and (cid:28)nd its properties in the transverse-tra eless (TT) gauge and also al ulate the geometry of the generated trapped surfa e [32, 33, 34, 35℄. In Se tion II we introdu e the prin ipal expressions [10℄ related to the Einstein-Rosen ylindri al GW [36℄ as, espe ially, represented in [10, 30℄. In Se tion III we dis uss the ylin- dri al GF (GW) in the ommutation number representation [8, 9℄ so that it may be thought of as omposed of a very large number of gravitational quanta whi h, like the quantum ones [8, 9℄, are reated and destroyed by the orresponding gravitational reation and destru - tion operators. In Se tion IV we relate the ylindri al lapse and shift fun tions [6, 13, 14℄ whi h determine the geometry of spa etime hypersurfa es to the mentioned gravitational reation and destru tion operators whi h, likewise, determine this geometry through reat- ing or (and) destroying the quantum omponents of the generating GW. We note that the ylindri al lapse and shift fun tions were related in [10℄ to the Einstein-Rosen parameters (see Eqs (29)-(30) in [10℄). The detailed al ulation relating the ylindri al lapse and shift A fun tions to the reation and destru tion operators are shown in Appendix . In Se tion V we show that, beginning from a gravitational ylindri al geometry in some hypersurfa e, THE EINSTEIN-ROSEN CYLINDRICAL GRAVITATIONAL WAVE 6 the probability to (cid:28)nd the same geometry upon this hypersurfa e tends to unity in the limit of the spa e Zeno e(cid:27)e t [25℄. As mentioned, we rather dis uss the spa e Zeno e(cid:27)e t and not the (intrinsi ) time analogue of it be ause in the ylindri al geometry one dis usses [10℄ the extrinsi time variable whi h is anoni aly onjugate to momentum [10℄ just as is any spatial variable. We note that it has been shown [37℄, using the examples of the quantum bubble and open-oyster pro esses [31, 38℄, that the mentioned stati and dynami quantum Zeno e(cid:27)e ts are also valid in quantum (cid:28)eld theory [31, 38℄. We also note in this respe t that the quantum Zeno e(cid:27)e t were dis ussed [39℄ in the framework of gravitomagnetism [40℄. B The detailed al ulations of the appropriate probability is shown in Appendix . In Se tion VI we graphi ally orroborate our theoreti al results so that one may see how the probabil- ity approa h unity in the Zeno limit. In Se tion VII we dis uss the ylindri al GW in the transverse-tra eless (TT) gauge whi h is hara terized by a very simpli(cid:28)ed formalism [13℄ in whi h, for example, the number of independent omponents of the related GW is mini- mal [13℄. In Se tion VIII we dis uss, using the method in [32℄, the related trapped surfa e [13, 32, 33, 34, 35℄ resulting from the passing ylindri al GW. In Se tion IX we summarize our dis ussion in a Con luding Remarks Se tion. 2 The Einstein-Rosen Cylindri al gravitational wave A spa etime is onsidered to be ylindri ally symmetri [10℄ if and only if one an show that (t, r, φ, z) −∞ < t < +∞ ∞ > r ≥ 0 2π > φ ≥ 0 there sxists a oordinate system , , , , −∞ < z < +∞ in whi h the line element be omes ds2 = −(N2 −e(ψ−γ)N2)dt2 +2N dtdr+e(γ−ψ)dr2 +R2e−γdφ2 +eγdz2, 1 1 (1) R ≥ 0 γ, ψ, N, N t r 1 where and are fun tions of and . The former dependen e of the g ,g ,g ,g ,g γ,ψ,R,N,N 11 22 33 00 01 1 nonzero metri tensor omponents upon the fun tions is, espe ially,designed[10℄tosuittheADM[6℄ anoni alformulationofgeneralrelativity. Thus, THE EINSTEIN-ROSEN CYLINDRICAL GRAVITATIONAL WAVE 7 N N 1 the and are, respe tively, the known ADM lapse and radial shift fun tions [10, 13℄. φ z The oordinates and are, essentially, (cid:28)xed ex ept for a possible trivial transformation of φ¯= ±φ+φ z¯= az+z t r 0 0 and whereas and may be subje t, without hanging the form of the line element from (1), to the more general transformation t¯= t¯(t,r), r¯= r¯(t,r) (2) (t, r, φ, z) One may show [10℄, using the Killing ve tor formalism in the system, that R γ the fun tions and are s alars. Thus, sin e the metri tensor oe(cid:30) ients depends, as t r (t,r) mentioned, only upon and one may write [10℄ the part of the line element (1) in the following onformally (cid:29)at form ds2 = e(γ¯−ψ)(−dt¯2 +dr¯2)+R2e−ψdφ2 +eψdz2, (3) R ψ φ z where the , , and are not barred due to their mentioned essential invarian y. Now, as emphasized in [10℄, if one writes the Einstein (cid:28)eld equations for the line element (3) one R ∂2R − ∂2R = 0 may realize that must be a harmoni fun tion whi h satis(cid:28)es ∂t¯2 ∂r¯2 . Thus, one R T may assume [10℄ to be a new radial oordinate and the time oordinate orresponding to it. That is, as emphasized in [10℄, the Einstein-Rosen oordinates an be uniquely and rigorously de(cid:28)ned by invariant pres riptions so that the line element (3) may be written as ds2 = e(Γ−ψ)(−dT2 +dR2)+R2e−ψdφ2 +eψdz2 (4) In su h ase the Einstein va uum equations are onsiderably simpli(cid:28)ed and redu e to the following three equations ∂2ψ ∂2ψ ∂ψ − −R−1 = 0 ∂T2 ∂R2 ∂R (5) THE EINSTEIN-ROSEN CYLINDRICAL GRAVITATIONAL WAVE 8 ∂Γ 1 ∂ψ ∂ψ = R ( )2 +( )2 ∂R 2 (cid:18) ∂T ∂R (cid:19) (6) ∂Γ ∂ψ ∂ψ = R ∂T ∂T ∂R (7) Asemphasized in[10℄, Eq(5)hasexa tly thesameformasthewave equationofthe ylin- ψ dri ally symmetri massless s alar (cid:28)eld advan ing in a Minkowskian spa etime whereas Eqs (6) and (7)are, respe tively, the energy density and the radialmomentum density of this (cid:28)eld in ylindri al oordinates. The solution of Eq (5) is obtained by using the separation of k variables method [30℄ so that the resulting wave fun tion for a parti ular wave number is ψ (R,T) = J (kR) A(k)e(ikT) +A∗(k)e−(ikT) , k 0 (8) (cid:0) (cid:1) j (kR) A(k), A∗(k) 0 where is the bessel fun tion of order zero [41℄ and are the amplitude and its omplex onjugate of the solution to the time part of Eq (5). Note that here we assume, c = h¯ = 1 w = k = p w, k, p as generally done in the relevant literature, that so that where k are respe tively the frequen y, wave number and momentum of some mode. Sin e is a ontinuous parameter one may obtain the general solution to Eq (5) by integrating over all k the modes . Thus, the relevant general wave fun tion is ∞ ψ(R,T) = dkJ (kR) A(k)e(ikT) +A∗(k)e−(ikT) 0 Z (9) 0 (cid:0) (cid:1) π (T,R) ψ The anoni al onjugate momentum may be obtained [30℄ by using the Hamilton equation ∂ψ = {ψ,H}, ∂t (10) ψ where is given by Eq (9), the urly bra kets at the right denote the Poisson bra kets and THE EINSTEIN-ROSEN CYLINDRICAL GRAVITATIONAL WAVE 9 H the Hamilton fun tion is ∞ H = dr N˜H˜ +N˜1H˜ 1 Z (cid:18) (cid:19) (11) 0 H˜ H˜ 1 The quantities and are respe tively the res aled superHamiltonian and supermomen- tum whi h where shown in [10℄ (see Eqs (93)-(97) and (106)-(108) in [10℄) to be 1 1 H˜ = R Π +T Π + R−1π2 + Rψ2 ,r T ,r R 2 ψ 2 ,r (12) H˜ = T Π +R Π +ψ π , 1 ,r T ,r R ,r ψ r Π , Π T R where the su(cid:30)xed apostroph denote di(cid:27)erentiation with respe t to and are the T R N˜ N˜ 1 respe tive momenta anoni ally onjugate to and . The quantities and respe tively N N π (T,R) 1 ψ denote the res aled lapse and shift fun tion , (see Eq (96) in [10℄). Thus, were shown [30℄ to have the form ∞ π (T,R) = iRR dkkJ (kR) A(k)e(ikT) −A∗(k)e−(ikT) ψ ,r 0 Z (13) 0 (cid:0) (cid:1) ∞ −RT dkkJ (kR) A(k)e(ikT) +A∗(k)e−(ikT) , ,r 1 Z 0 (cid:0) (cid:1) j (kR) 1 where is the (cid:28)rst order Bessel fun tion [41℄ whi h may be obtained by di(cid:27)erentiating j (kR) R j (kR) = −kj (kR) ψ(T,R) π (T,R) 0 0 ,R 1 ψ with respe t to as . The initialdata for and T = 0 R = r Q(r) P(r) are al ulatedfor and andare, respe tively, denotedby and asfollows ∞ Q(r) = ψ (r) = ψ(T,R)| = dkJ (kr) A(k)+A∗(k) 0 T=0,R=r 0 Z (14) 0 (cid:0) (cid:1) ∞ P(r) = π (r) = π (T,R)| = ir dkkJ (kr) A(k)−A∗(k) ψ0 ψ T=0,R=r Z 0 (15) 0 (cid:0) (cid:1) THE EINSTEIN-ROSEN CYLINDRICAL GRAVITATIONAL WAVE 10 A(k) A∗(k) Solving the last two equations for and one obtains 1 ∞ A(k) = drJ (kr) krQ(r)−iP(r) 0 2 Z (16) 0 (cid:0) (cid:1) 1 ∞ A∗(k) = drJ (kr) krQ(r)+iP(r) 0 2 Z (17) 0 (cid:0) (cid:1) A(k), A∗(k) One may show, using Eqs (16)-(17), that the variables satisfy the following Poisson bra kets ∞ δ(A(k))δ(A∗(k′)) δ(A(k))δ(A∗(k)) iδ(k −k′) {A(k),A∗(k′)} = dr − = Z (cid:20)δ(Q(r)) δ(P(r)) δ(P(r)) δ(Q(r)) (cid:21) 2 0 {A(k),A(k′)} = {A∗(k),A∗(k′)} = 0, (18) where use was made of the relation [30℄ ∞ 1 drrj (kr)j (k′r) = δ(k −k′), n = 0,1,2,3.... n n Z k′ 0 Q(r) In a similar manner, using Eqs (14)-(15), it is possible to show that the variables and P(r) satisfy the following Poisson bra kets ∞ δ(Q(r)) δ(P(r′)) δ(Q(r)) δ(P(r′)) {Q(r),P(r′)} = dk − = −2iδ(r −r′) Z (cid:20)δ(A(k))δ(A∗(k)) δ(A∗(k)) δ(A(k))(cid:21) 0 {Q(r),Q(r′)} = {P(r),P(r′)} = 0, (19) where use was made of the relation [30℄ ∞ δ(r −r′) dkkJ (kr)J (kr′) = , n = 0, 1,2,... n n Z r′ 0 Using Eqs (6), (9) and the unnumbered relation written just after Eq (18) one may obtain