• Nenhum resultado encontrado

Braz. J. Phys. vol.31 número2

N/A
N/A
Protected

Academic year: 2018

Share "Braz. J. Phys. vol.31 número2"

Copied!
10
0
0

Texto

(1)

Critial Phenomena in Gravitational

Collapse: The Studies So Far

Anzhong Wang

Departamento deFsiaTeoria,

UniversidadedoEstadodoRiodeJaneiro,

RuaS~aoFranisoXavier524,Maraan~a, 20550-013,RiodeJaneiro,RJ,Brazil

Reeivedon22February,2001

Studies of blak hole formation from gravitational ollapse have revealed interesting non-linear

phenomenaat thethresholdofblakhole formation. Inpartiular,in1993 Choptuikstudiedthe

ollapseofamasslesssalareldwithspherialsymmetryandfoundsomebehaviour,whihisquite

similartotheritial phenomenawell-knowninStatistialMehanisandQuantum FieldTheory.

Universality and ehoing of the ritial solutionand power-law saling of the blak hole masses

havegivenrisetothenameCritialPhenomenainGravitationalCollapse. Choptuik'sresultswere

soononrmedbothnumeriallyandsemi-analytially,andhaveextendedtovariousothermatter

elds. Inthispaper,weshallgiveabriefintrodution tothisfasinatingandrelatively newarea,

andprovideanupdatedpubliationlist. Ananalytial\toy"modelofritialollapseispresented,

andsomeurrentinvestigationsaregiven.

I Introdution

Gravitationalollapseofarealistibodyhasbeenone

ofthe mostimportantandthornysubjets inGeneral

Relativity (GR) sine the veryearly times of GR [1℄.

The ollapse generallyhas four kinds of possible nal

states. The rstis simplythe haltof theproess in a

self-sustainedobjet,suhas,stars. Theseondisthe

dispersion of the ollapsing objet and nally leaves

behind a at spaetime. The third is the formation

of blak holeswith outgoinggravitationaland matter

radiation, while the fourth is the formation of naked

singularities. For the last ase, however, the osmi

ensorshiphypothesis[2℄delaresthatthesenaked

sin-gularitiesdonotourin Nature.

DuetothemathematialomplexityoftheEinstein

eldequations,wearefrequentlyforedtoimposesome

symmetriesonthe onernedsystemin orderto make

theproblemtratable. Spaetimeswithspherial

sym-metryareoneoftheases. Inpartiular,gravitational

ollapseofaminimally oupledmasslesssalareld in

suh spaetimes wasstudied both analytially[3℄ and

numerially[4℄, and somefundamental theorems were

established. Quitereentlythisproblemhasfurther

at-tratedattention, due to Choptuik'sdisoveryof

rit-ialphenomena that were hitherto unknown[5℄. As a

matteroffat,itissoattrativethatCritial

Phenom-enain GravitationalCollapse hasalreadybeenavery

establishedsub-areainGR,andseveralomprehensive

reviewartilesalreadyexist[6, 7,8,9,10,11, 12, 13℄.

Inthispaper,asummaryofaninvitedtalkgivenat

the XXI Brazilian National Meeting on Partiles and

Fields,weshallrstbrieyreviewthesubjetandgive

an updatedlist of publiation in this area,whih will

bedonein Se. II.Then, in Se. III,weshallpresent

ananalyti\toy"modelofaollapsingmasslesssalar

eld. Theword\toy"modelheremeansthatthemodel

doesn'treally representritialollapse, sinethe

per-turbationsoftheorresponding\ritial"solutionhave

morethan one unstablemode. However, itdoeshave

allthemainfeaturesofritialollapse. Sinesofar,no

anyritialsolutionisknownexpliitlyinaloseform,

thistoymodelstillservesasagoodillustrationto

rit-ialphenomenaingravitationalollapse. Thepaperis

losedbySe. IV,inwhihsomeurrentinvestigations

in thisfasinatingareaaregiven.

II Critial phenomena in

gravi-tational ollapse

Startingwithspherialspaetimes,

ds 2

=

2

(t;r)dt 2

+a 2

(t;r)dr 2

+r 2

d 2

; (1)

where d d 2

+sin 2

d' 2

, and fx

g = ft;r;;'g

aretheusualspherialoordinates,Choptuik[5℄

inves-tigatedgravitationalollapseofamasslesssalareld,

(2)

,whihsatisestheEinstein-salareldequations, R = ; ; ;

2 = 0; (2)

whereR

denotestheRiitensor,[8G= 4

℄isthe

gravitationalouplingonstant,()

;

=()=x ; 2 g r r

, and r

denotes the ovariant derivative.

One an initial smooth onguration of the massless

salareldisgiven,theseequationsuniquelydetermine

thelaterevolutionofthespaetimeandthesalareld

[3℄. Let the initial distribution of the massless salar

eldbeparameterizedsmoothlybyaparameterpthat

haraterizesthestrengthoftheinitialonditions,suh

thattheollapseofthesalareldwiththeinitialdata

p >p

forms ablakhole, while the onewith p<p

doesnot. Asimpleexampleisthegaussiandistribution

ofthemasslesssalareld

(t

0 ;r)=

0 r r 0 3 exp r r 0 Æ q ; (3) where t 0

denotes the initial time of the ollapse, and

0 ; r

0

; Æ,and qareonstants[SeeFig. 1℄.

φ

r

0

r

φ

( t , r)

0

0

0

δ

Figure 1. The initial onguration of the massless salar

eld att=t

0

givenby Eq.(3)inthe text. It atually

rep-resents a ollapsing spherial shell, made of the massless

salar eld,withitsthiknessÆandentralizingatthe

ra-diusr=r

0 .

0

representstheamplitudeofthewavepaket.

Inthisase,Choptuik foundthattheparameterpan

beanyofthefourparameters,

p=f

0 ; r

0

; q; Æg; (4)

that is,xinganythree ofthefour parameters,for

ex-ample,r

0

; Æandq,andleavingonly

0

hange,weshall

obtain a familyof initial data, S[

0

℄. For this family

of initialdata Choptuikfoundthat there existsa

rit-ial value

0

suh that when

0 >

0

theollapse

al-waysformsblakholes,andwhen

0 <

0

themassless

salar eld rst ollapses, then disperses to spaelike

innity,andnallyleavesaatspaetimebehind

with-outforminganykindofspaetimesingularities. When

0 0 + 0 , where 0

is verysmall,after ertain

theritial one (

0 =

0

). But, at the veryend, the

ollapsewillsuddenlyruns awayfrom theritialone,

by either forming blak holesordispersingto innity,

dependingonthesignsof

0

. Choptuikfoundthatfor

theongurationofEq.(3)therearefourdierent

fam-ilies ofinitial data, S[

0 ℄; S[r

0

℄; S[Æ℄ and S[q℄, whih

allexhibit theabovebehaviour.

In addition to these four, Choptuik also studied

many others and found that for all the families that

behaveasabove,the so-alledgenerismooth families

of initial data, all the ritial solutions are idential,

or in another word, universal. Moreover, the ritial

solutionisalsoperiodi, thatis,

A

(;)=A

(; +4); (5)

whereA =f ;a ; g,and =ln t r 0

; =ln r t 0

(); (6)

withr

0

beingadimensionfulonstant,and

0

()a

pe-riodiotherwisearbitraryfuntionwithperiod4. The

onstant4isadimensionless onstant,whihwas

nu-meriallydeterminedas43:447.

Yet,near theritialsolutionbut withp>p

, the

massofblakholestakesthesalingform

M

BH

=K(p p

)

; (7)

where K is a family-dependent onstant, but is

an-otherdimensionless universalonstant,whihwas

nu-meriallydeterminedas0:37.

Universalityandehoingoftheritialsolutionand

power-lawsaling of theblakhole masseshave given

riseto the name Critial Phenomena in Gravitational

Collapse.

Choptuik's results were soon onrmed by several

independent studies both numerial [14℄ and

semi-analytial[15℄,andhavebeenextendedtoothermatter

elds,suhas,

Axisymmetrigravitationalwaves[16℄;

Perfetuids with theequation of statep=k,

where p denotes the pressure of the uid and

theenergydensity,andkisaonstant[17,18℄;

Quantum blak holeformation in 2-dimensional

spaetimes[19℄;

Non-linear -models in two dimensional target

spae[20℄;

Masslesssalareldin Brans-Diketheory[21℄;

SU(2)Yang-Millseld [22℄;

(3)

Gravitationallyollapsingprimordialdensity

u-tuations intheradiationdominatedphaseofthe

earlyUniverse [25℄;

SU(2)Skyrmeeld [26℄;

The ollapse of ollisionless matter of the

Einstein-Vlasovequations[27, 28℄;

Topologial domain walls interating with blak

holes[29℄;

Thegravitationalollapseofmasslesssalareld

in higherdimensional spaetimes[30℄;

Non-linear -models in three dimensional target

spae[31℄;

GravitationalollapseinTensor-Multi-Salarand

Non-linearGravityTheories[32℄;

Bosonstars[33℄;

Massless salar eld oupled with the

osmo-logialonstantin (2+1)-dimensionalspaetimes

[34℄.

In review of all these studies, nowthe followingis

lear:

(a) In general the ritial solution and the two

di-mensionlessonstants4andareuniversalonly

withrespettothesamemattereld,andusually

are matter-dependent. Forexample, forthe

ol-lapseoftheSU(2)Yang-Mills eld,itwasfound

[22℄ that 4 0:74 and 0:2, while in the

aseofmasslesssalareld,Choptuikfoundthat

43:447and0:37.

(b) The ritial solutions an have disrete

self-similarity(DSS)[35℄orontinuousself-similarity

(CSS) [36℄, or none of them, depending on the

mattereldsandregionsoftheinitialdataspae.

So far, in all the ases where the ritial

solu-tioneither hasDSS orCSS,blakholesform

al-ways startingwith zeromass,and taketheform

ofEq.(7),theso-alledTypeII ollapse,whilein

theasesinwhihtheritialsolutionhasneither

DSSnorCSS,theformationalwaysturnsonwith

amass gap,theso-alled TypeI ollapse,

orre-sponding, respetively, to the seond- and

rst-order phase transitions in Statistial Mehanis

[37℄.

() The universality of the ritial solution and the

exponent now arewell understood in termsof

perturbations ofritial solutions[18℄, whilethe

one of 4 still remains somewhat of a mystery.

Theformeris loselyrelatedtothefat thatthe

perturbationsoftheritialsolutionhasonlyone

unstable mode. This propertynowis onsidered

asthemain riterionforasolutionto beritial

Tounderstandthe lastpropertybetter,letus

on-sider the phase spae, that is, onsider GR as an

innite-dimensional dynami system. If we make a

(3+1) split of the spaetime, for example, following

the Arnowitt, Deser, and Misner (ADM)

deomposi-tion,wewillndthatthedynamiquantitieswillbethe

indued spatial three metri, the extrinsi urvature,

and the matter distribution. Then, the phase spae

will onsist of all the possible three metris, extrinsi

urvature, andongurationsofthematterelds. For

the aseof masslesssalareld,from theno-hair

the-oremofblakholes[38℄,weknowthat theonlystable

blakholesolutionoftheEinstein-salareldequations

is the Shwarzshild blak holewith aonstant

mass-less salar eld. Exept for this blak hole, another

stable stateis theMinkowskianspaetime. Ofourse,

wealsoknowthattheollapseofamasslesssalareld

an form naked singularities,too, but so farwedon't

know if they are stable or not [1℄. At this point, we

shall adopt thepointof viewof theosmi ensorship

onjeture [2℄, and assume that they are not stable.

Otherwise, theremayexist twomoreritialsolutions

thatseparate,respetively,blakholesfromnaked

sin-gularities,andatspaetimesfromnakedsingularities.

However,thisdoesn'taetourfollowingdisussionsif

weare restrited onlyto theboundary betweenblak

holesandatspaetimes,andtheanalysisanbeeasily

extended tootherboundaries.

p*

p > p*

*

A

S

S[p]

Flat

Space-Times

Black

Holes

Figure 2. The phase spae of the dynami system of the

Einstein-salar eld equations. Thehypersurfae S is the

ritial surfae of odimension one, whih separates the

basin of blak holes from the basinof at spaetimes. A

generismoothfamilyofinitialdataS[p℄alwayspassesthe

twobasins at the ritial point p = p

onthe ritial

hy-persurfae. All the initial data on the hypersurfae will

ollapseto theritial solutionA

that isaxedpoint on

the hypersurfae whenit has CSS anda xed ylewhen

it hasDSS.Alldetails ofinitial dataare soonwashedout

during theollapsingproess,andthe ollapsewithinitial

dataneartheritialpointwillbeverysimilartothe

riti-alollapse. Thissimilarityanbelastalmosttothexed

point A

,whereby theoneunstablemodesuddenlydraws

the ollapse eitherto form blak hole or a atspaetime,

dependingonwhetherp>p

orp<p

.

(4)

theatspaetimeswithdierentonstantvaluesofthe

salareld,areattrativexedpoints,whileinsidethe

basinofblakholes,theShwarzshildblakholeswith

dierent masses are the attrating xed points. The

boundary between the two attrative basins is alled

the ritial surfae, and the ritial solution always

lies on it. Sine it has only one unstable mode, this

surfae must be a hypersurfae of odimension one,

that is, one dimension less than the original

innite-dimensional phasespae. Bydenition, aphase spae

trajetoryneverleavesthishypersurfae,ifitisinitially

onit,butapproahestotheritialsolution,whihisa

xed point onthishypersurfaeifthe ritialsolution

isontinuousself-similar,oraxedyleiftheritial

solution is disrete self-similar [13℄. Within the

om-plete phase spae, the ritial solutionis anattrator

of odimension one, i.e., it has an innite number of

deayingperturbation modestangentialtothe ritial

hypersurfaeandasinglegrowingmodeperpendiular

tothehypersurfae. Anytrajetorybeginningnearthe

ritialhypersurfae,butnotneessarilynearthe

riti-alpoint(oryleintheDSSase),rstmovesparallel

to thehypersurfaeand goesdownalmost to the

rit-ial point (or yle), then is suddenly drawn away by

thesingleunstablemodeintheperpendiulardiretion,

andnallyendsupatoneofthexedpoints,byeither

formingablakholeoraatspaetime. Duringthe

dy-namiproess,alldetailsoftheinitialdataarequikly

washed away, exept for the distane from the blak

hole threshold. Therefore, for the both super-ritial

(p>p

)andsub-ritial(p<p

)ollapse,thereexists

adomain,p

4ppp

+4p,inthephasespae,

in whihtheollapseisverysimilar totheritialone

during ertainperiodoftimes[SeeFig.2℄.

III Critial ollapse of massless

salar eld: an analyti Toy

model

Inthissetion,weshallpresentalassofanalyti

solu-tionsoftheEinstein-salareldequations,whih

repre-sentsgravitationalollapseofawavepaketonsisting

ofmasslesssalareld[39℄. Thislassofsolutionswas

rst disovered by Roberts [40℄ and later studied by

severalauthors in the ontext of ritialollapse [41℄.

Asweshallshowbelow,thesesolutionspossessmostof

thefeaturesofritialphenomena,althoughtheydon't

exatlyrepresentritialollapse,beausethesolution

thatseparatestheformationofblakholesfromthatof

atspaetimeshasmorethanoneunstablemode[42℄.

It is exatly in this sense, we refer these solutions as

representinga\toy"modelofritialollapse.

TheRobertssolutionsaregivenby[40℄

ds 2

= G(u;v)dudv+r 2

(u;v)d 2

; (8)

whereuandv representtwonulloordinates,interms

ofwhihthe metrioeÆientsandthe orresponding

masslesssalareld are given,respetively,by

r(u;v) = 1

2 u

2

2uv+4b

2 v

2

1=2

;

G(u;v) = 1; (9)

(u;v) = 1

p

2 ln

(u v) (1 4b

2 )

1=2

v

(u v)+(1 4b

2 )

1=2

v

;(10)

where b

2

is an arbitrary onstant. Note that the

no-tations used here losely follow the ones used in [39℄

butslightlydierentfrom theones used in[40℄. From

Eq.(9)itanbeeasilyshownthat theloalmass

fun-tion[43℄ isgivenby

m(u;v) r

2 1 r

; r

; g

=

(1 4b

2 )uv

8r

; (11)

whih is zero onthe hypersurfaev =0 and negative

for u; v < 0. Thus, to have a physially reasonable

spaetimeweneedto restrittheabovesolutionsvalid

onlyin theregionu0; v0. Sinethemassiszero

onthehypersurfaev=0,wemayjointheabove

solu-tionsarossthehypersurfaev=0withaMinkowskian

spaetime. Asshownin[39℄,thisispossibleifthe

met-riintheregionv0takestheformofEq.(8)butwith

themetrioeÆientsandthemasslesssalareld

be-inggivenby

r(u;v) = a(v) 1

2

u a(0);

G(u;r) = 2a 0

(v); =0; (v<0); (12)

wherea(v)isanarbitraryfuntionsubjettoa 0

(v)>0

and a 0

(0) = 1=2, and a prime denotes the ordinary

dierentiationwithrespet totheindiatedargument.

Forsuha mathing, itanbeshownthat the

hyper-surfae v = 0 is free of any kind of matter and

rep-resentsa boundary surfae [44℄. The region v < 0 is

Minkowskian[SeeFig.3℄.

Ontheotherhand,fromEqs.(9)and(10)itanbe

alsoshownthatthespaetimeintheregionu<0;v>0

represents a ollapsing massless salar wave. When

b

2

<0, thesalarwaveollapsesinto aspaetime

sin-gularityonthehypersurfaeu= [(1 4b

2 )

1=2

1℄v,

whihispreeded byanapparenthorizonat u=4b

2 v.

Thus,theorrespondingsolutionsrepresentthe

forma-tionof blak holes. When b

2

= 0, the singularity

o-inides with the apparent horizon on the null

hyper-surfaeu =0. When0 < b

2

<1=4, it an beshown

that the massless salar eld rst ollapses and then

dispersesintoinnity,withoutformingblakholes,but

instead,leavesaMinkowskianspaetimebehind,whih

nowisrepresentedbytheregion,u;v>0,inwhihthe

metritakestheform ofEq.(8),butwith

G(u;v) = 4b 1=2

b 0

(u); r=b 1=2

(5)

(u;v) = 1 p 2 ln

1+(1 4b

2 )

1=2

1 (1 4b

2 )

1=2

;

(0<b

2

<1=4; u;v>0); (13)

where b(u)is anarbitraryfuntion,subjetto b 0

(u)>

0;b 0

(0)=1=(4b 1=2

2

). Oneanshow that the

hypersur-faeu=0;v>0isalsofreeofanykindofmatterand

representsaboundarysurfae.

Vaidya

u

v

0

v

φ

M

4

0

Figure 3. The spaetime inthe (u; v)-plane. Theregion

v<0isMinkowskian,whiletheregionu0; v0

repre-sentsgravitationalollapseofthemassless salareld. (a)

When b2 < 0, the salar wave ollapses into a spaetime

singularity at u = [(1 4b

2 )

1=2

1℄v <0, whihis

pre-ededby anapparenthorizon loatedat u= 4jb

2 jv <0.

(b)Whenb

2

=0,the spaetimesingularity oinideswith

the apparent horizon on u = 0 whih is null. () When

0 <b2 < 1=4, the massless salar eld rst ollapses and

thendispersesintoinnity,andnallyleavesaMinkowskian

spaetimebehindintheregionu; v>0.

Inthe ase b

2

< 0, where blak holes are formed,

Eq.(11)showsthatontheapparenthorizonu= 4jb

2 jv

themassbeomesunboundedasv!+1. Inorderto

haveblak holeswith nite mass, we shall follow [39℄

rsttoutthespaetimealongthehypersurfaev=v

0

and then join the region 0 v v

0

with an

asymp-totiallyatregion. Tomodeltheout-goingradiation

of themasslesssalareld, weshall hoose theregion

vv

0

asdesribedbytheVaidyasolution[45℄,

ds 2 = 1 2m(U) r dU 2

+2dUdr+r 2

d 2

;

(vv

0

); (14)

where U is the Eddington retarded time, whih is in

generalthefuntionofuappearinginEq.(8),andm(U)

istheloalmassoftheout-goingVaidyadust. The

or-respondingenergy-momentumtensorisgivenby

T + = 2 r 2 dm(U) dU Æ U Æ U

; (vv

0

): (15)

The hypersurfae v = v

0

in the oordinates fx

g =

fU; r; ; 'gisgivenby

dU(r)

=

2r

; (vv

0

): (16)

Then, itanbeshown thatthejuntion onditionson

thehypersurfaev=v

0

require

M(r) m(U)j

v=v0 = 1 r h p(4p 2 +r 2 ) 1=2 2p 2 i ; v 0 = 4p 2 p ; (17)

wherepistheintegrationonstant,and

p (1 4b 2 ) 1=2 4 v 0 : (18)

Forthedetails, we referreaders to [39℄. Sine for the

abovemathing,thehypersurfaev=v

0

isfreeof

mat-ter,thefuntion M(r)representsthetotalmassofthe

ollapsing wavepaket lledin theregion 0v v

0 .

Atthepastnullinnity,Eq.(17)showsthat

M(r!+1)=p; (19)

that is, theparameterpinthepresentaserepresents

thetotalinitialmassofthemasslesssalarwavepaket

withwhihitstartstoollapse.

As r ! 0 +

, from Eq.(17) we an see that M(r)

behavesas

M(r)! (

+1; p>p

,

0; p=p

,

1; p<p

.

(20)

M (r)

(a)

f = r/2

(b)

(c)

0

p

r

AH

r

*

Figure 4. The mass M(r) of ollapsing spherial shell.

The line (a) orresponds to the ase where p > p

, in

whih a blak hole is formed, and its mass is given by

MBH = M(rAH). The line (b) orresponds to the ase

where p = p

, while the line () orresponds to the ase

wherep<p

.

On the other hand, it is well-known that the

ap-parent horizon at r =2M(r)of the out-goingVaidya

(6)

Thus, by omparing the mass M(r) with r=2 we an

tellwhethertheollapseformsablakholeornot,

M(r) r

2 =

(

4p

2

+r 2

4r 2

[p+(4p

2

+r 2

)℄ )

1=2

h

4(p 2

p

2

) r 2

i

: (21)

Clearly,onlywhenp>p

,thesalareldandthenull

shell willollapseinside theeventhorizonat

r

AH =2(p

2

p

2

) 1=2

: (22)

r = 0

A H

EH

φ

Vaid.

M

0

v

0

4

(a)

M

0

v

r = 0

Vaid.

4

0

φ

(b)

M

M

0

v

4

0

4

Vaid.

0

φ

(c)

Figure 5. The orresponding Penrose diagrams. (a) The

asewherep>p

,inwhihablakholeis formed,andits

mass is givenby MBH = M(rAH). (b)It orrespondsto

thease where p=p

,inwhih thespaetime singularity

beomesnull andoinideswiththeapparent horizon. ()

Itorrespondstotheasewherep<p

,inwhihno

spae-timesingularityisformed,instead,whenthewavepaketof

themasslesssalareldollapsestothehypersurfaeu=0,

allofitsmasshasbeenradiatedawayandnothingisleftto

ollapse, so thespaetime intheregion u; v0beomes

Minkowskian.

Whenp=p

,M(r)=r=2ispossibleonlyattheorigin,

r=0,where azero-masssingularityis formed. Thus,

thesolutionwith p=p

representsthe\ritial"

solu-tionthat separatesthesuperritialsolutions(p>p

)

from the subritial ones (p < p

). In the subritial

ase, M(r) is always less than r=2, and the ollapse

neverforms ablakhole[f. Fig.4℄.

In the subritial ase, the region u;v > 0should

be replaed by the Minkowskian solution (13). As

shown above, the mathing aross the hypersurfae

u=0; 0vv

0

issmooth,i.e.,nomatterappearson

it. Toshowthatitisalsotheaseonthehypersurfae

u=0; vv

0

,whihseparatestheVaidyasolution(14)

from theMinkowskianone(13),werst makethe

o-ordinatetransformationU =U(u),andthen writethe

metri(14)intermsofu. Usingtheresultsobtainedin

[46℄,oneanshowthatto haveasmoothmathingwe

havetoimpose theondition

U 0

(0)= 1

(4b

2 )

1=2

; (23)

(7)

Clearly,byproperlyhoosingthe funtion-dependene

of U(u), the rst ondition (23) an be always

satis-ed. Ontheotherhand,fromEqs.(9)and(17)onean

showthatthelastonditionisalsosatisedidentially.

Therefore, themathing of the Vaidyasolutionto the

Minkowskianonearossthehypersurfaeu=0; vv

0

isalwayspossibleforp<p

. Theorresponding

Pen-rose diagram for eah of the three asesare shown in

Fig. 5.

Ontheapparenthorizonr=r

AH

,thetotalmassof

thesalarwavepaket isgivenby

M

BH =

1

2 r

AH

=K(p p

) 1=2

; (25)

where K (p+p

) 1=2

. The above expression shows

that theblakholemass takesapower-lawform with

its exponent =0:5, whih is dierentfrom theDSS

ase, where Choptuik found 0:37. As we

men-tionedpreviously,theabovesolutionsdon'treally

rep-resent ritial ollapse, beause the \ritial" solution

given above has more than one unstable modes [42℄.

Therefore, the dierent value of obtained here does

not means any ontridition to Choptuik's numerial

results.

IV Current Investigations

As Critial Phenomena in gravitational ollapse is a

rathernewareainGR,therearemanyopenproblems.

Inthefollowingweshallmentionsomeofthem.

A. The eets of angular momentumin ritial

ollapse

As we know, the angular momentum plays avery

signiantroleinblakholephysis,andallthe

realis-tibodies,suhas,neutronstars,havenon-zeroangular

momentum. Thus,itisveryimportanttostudythe

ef-fetsofangularmomentumonritialollapse. Sofar,

all the studies of ritial phenomena in gravitational

ollapse havebeenrestrited to spherialase, exept

for the works of Abrahams and Evans [16℄ and

Alu-bierreet al. [47℄. In[16℄ the authors studied the

ol-lapseof axisymmetri purely gravitationalwaves,and

found the type II ritial ollapse. However, in this

studythetotalangularmomentumisstillzero. In[47℄

theollapseofpureBrilltypegravitationalwavesin3D

Numerial Relativity wasstudied and the ritial

am-plitudeforblakholeformationwasdetermined.

How-ever,duetotheomplexityoftheproblem,nosuÆient

evideneforritialollapsewasobserved.

In addition to the above, Gundlah and his

o-workers [48℄ studied the problem using non-spherial

perturbations, and in partiular found that all the

modes of non-spherial perturbations of the massless

salar eld studied initially by Chpotuik are stable,

the spherial ase may remain ritial even in

non-spherial ase. He also found that small angular

mo-mentumalsotakesasalingformneartheritialpoint

but with adierentexponent. Besides, Rein, Rendall

andShaeer[27℄studiedthespherialollapseof

ol-lisionlessmatterthatonsistsofounter-rotating

parti-les,andfoundthatonlyTypeIritialollapse. This

result was further onrmed by Olabarrieta quite

re-ently[28℄.

Movingfromspheriallysymmetriaseto

axisym-metrione,theproblembeomesmuhmathematially

involved, and very sophistiated (numerial)

meth-ods are needed. Choptuik, Hirshmann and Liebling,

among others, havebeenworking on this problem

re-ently [49℄, and are expeted to report their results

soon.

B. The quantumeets on ritialollapse

Critialphenomena areatuallyphenomena in the

stronggravitationaleld regime,and Quantum eets

should be very important for the formation of blak

holeswithverysmallmass. ChibaandSiino[50℄

stud-iedthisproblem andshowedthattheQuantumeets

maydestroythetypeIIritialphenomena,whileAyal

andPiran[51℄ showedthattheydon't,butrathershift

theritialvaluep

. However,sineinbothofthetwo

asesthe Quantum eetive energy-momentumtensor

(EMT)wastakenfromtwo-dimensionaltoymodel,the

onsisteneofsuhanEMTwiththefour-dimensional

gravitational ollapse is still an open question.

Re-ently,Bradyand Ottewill[52℄alulatedtheeetive

EMTofaonformallyoupledsalareld onthexed

bakgroundoftheritialsolutionsoftheperfetuid

with the equationof state p=k in four-dimensional

spherial spaetimes, and found that when k < 0:53,

the Quantum eets destroy the type II ritial

phe-nomena, whilewhen k >0:53 theiralulationsbreak

down,andadenitiveonlusionisstill absent.

C.Theappliationofrenormalizationgroup

the-ory to ritialollapse

The Renormalization Group Theory has ahieved

great suess in the studies of ritial phenomena in

StatistialMehanis[37℄,andseveralauthors,

inlud-ingArgyres[53℄,andKoike,HaraandAdahi[18℄,have

pointedoutthatthetimeevolutionneartheritial

so-lution ingravitationalollapsemayalso beonsidered

as arenormalizationgroupow onthe phasespaeof

initial data. As a matter of fat, the analysis of the

phasespaegivenin theIntrodutionexatlyfollowed

this idea. However, this analysis is valid onlyin

self-similar spaetimes. Inorder toobtainafull

renormal-ization group, one needs to generalize them to

arbi-traryspaetimes,whih isturnedoutnottrivial,as in

(8)

ar-initial stepsto thisdiretion,but asuessful

applia-tion of the Renormalization Group Theory to ritial

ollapsestillremainsasanopenquestion.

Besides the above mentioned problems many

oth-ers are also under the urrent investigations, suh as,

ndingmoremattereldsthatexhibitritialollapse,

inludinguniversallasses;applyingtheanalysisof

per-turbations of blak holes to ritial solutions;

under-standingitsphysialoriginoftheonstant4[55℄;

nd-ing somepossibleastrophysialobservationsof ritial

phenomena,andsoon. Inpartiular,itwasknownfor

alongtime that theollapseof neutronstarsexhibits

thetypeIritialphenomena[56℄. Animportant

ques-tionisthat: DoesthistypeIritialollapsehaveany

observationalonsequene?

Forfurtherreferenesofritialphenomenain

grav-itationalollapse,wewouldliketoreferthereadersto

thereviewartiles[6,7,8,9,10,11,12, 13℄.

Aknowledgments

The author would like to express his gratitude to

M.W. Choptuik, S.L. Liebling, J. Pullin, and W.M.

Suenforvaluabledisussionsandsuggestionsinritial

ollapse. Hewouldalsolikeverymuhtothankhis

ol-laboratorsinthisarea,C.F.C.Brandt,R.-G.Cai,E.W.

Hirshmann,L.-M.Lin,H.P.deOliveira,J.F.Villasda

Roha, N.O. Santos, and Y.M. Wu. Thenanial

as-sistanefromCNPqis gratefullyaknowledged.

Referenes

[1℄ P.S.Joshi, GlobalAspets inGravitationand

Cosmol-ogy (Clarendon Press, Oxford, 1993); \Gravitational

Collapse: TheStorysofar,"gr-q/0006101.

[2℄ R.Penrose,Riv.NuovoCimento1,252(1969).

[3℄ D. Christodoulou, Commun. Math. Phys. 105, 337

(1986);ibid.106,587(1986);ibid.109,591(1987);ibid.

109,613(1987);Commun.PureAppl.math.XLIV,339

(1991); Ann. Math. 140, 607 (1994); ibid., 149, 183

(1999).

[4℄ D.S. Goldwirth and T. Piran, Phys. Rev. D36, 3575

(1987).

[5℄ M.W.Choptuik,\CritialBehaviourinMasslessSalar

Field Collapse," inApprohes to NumerialRelativity,

Proeedings of the International Workshop on

Numer-ial Relativity, Southampton, Deember, 1991, Edited

byRayd'Inverno;Phys.Rev.Lett.70,9(1993);M.W.

Choptuik,\CritialBehaviourinSalarFieldCollpase,"

inDeterministiChaosinGeneralRelativity,Editedby

D. Hobill etal. (Plenum Press, New York, 1994), pp.

155-175;M.W.Choptuik,E.W.Hirshmann,andS.L.

Liebling,Phys.Rev.D55,6014 (1997).

[6℄ A.M.AbrahamsandC.R.Evans,Gen.Relat.Grav.26,

379(1994).

[7℄ C. Gundlah, \Critial phenomena in gravitational

143, Institute ofMathematis, PolishAademy of

Si-ene (Banah Center Publiations), Warszawa 1997,

gr-q/9606023.

[8℄ P. Bizon, Ata Cosmologia 22, 81 (1996),

gr-q/9606060.

[9℄ J.Horne,\Critialbehaviorinblakholeollapse,"

Mat-tersofGravity7,14(1996),gr-q/9602001.

[10℄ M. W. Choptuik, \The (unstable) threshold of blak

hole formation," talk given at GR15, Pune, India, to

appearintheproeedings,gr-q/9803075.

[11℄ C. Gundlah, Adv. Theor. Math. Phys. 2 1 (1998),

gr-q/9712084.

[12℄ S.L. Liebling, J. Phys. PRAMANA (India), 55, 497

(2000),gr-q/0006005.

[13℄ C.Gundlah,\Critialphenomenaingravitational

ol-lapse,"gr-q/0001046.

[14℄ C.Gundlah,J.Pullin,andR.Prie, Phys.Rev.D49,

890(1994);D.Garnkle, Phys.Rev.D51,5558(1995);

R.S.HamadeandJ.M.Stewart,Class.QuantumGrav.

13,497 (1996), S.Hod and T.Piran,Phys.Rev.D55,

R440(1997).

[15℄ C.Gundlah,Phys.Rev.Lett.75,3214(1995); Phys.

Rev.D55,695(1997).

[16℄ A.M.AbrahamsandC.R.Evans,Phys.Rev.Lett.70,

2980(1993);Phys.Rev.D49,3998(1994).

[17℄ C.R.Evansand J.S.Coleman,Phys.Rev.Lett. 72,

1782(1994); D. Maison, Phys.Lett. B366,82 (1996);

M.Goliath,U.S.Nilsson,andC.Uggla,Class.Quantum

Grav. 15,2841 (1998);D.W. Neilsen andM.W.

Chop-tuik,ibid., 17, 733(2000); ibid., 17,761 (2000); B.J.

Carr,A.A.Coley,M.Goliath,U.S.Nilsson,andC.

Ug-gla,Phys.Rev.D61,081502(2000);C.Gundlah,

\Crit-ialgravitational ollapseofa perfetuidwithp=k:

Nonspherialperturbations,"gr-q/9906124.

[18℄ T.Koike,T.Hara,andS.Adahi,Phys.Rev.Lett.74,

5170(1995);Phys.Rev.D59,104008(1999).

[19℄ A.StromingerandL.Thorlaius,Phys.Rev.Lett.72,

1584 (1994); J.-G. Zhou, H.J.W. Muller-Kirsten, and

M.-Z.Yang,Phys.Rev.D51,R314(1995);Y.Peleg,S.

Bose,andL.Parker,ibid.,D55,R4525(1997);L.Parker

andY.Peleg,\QuantumMassGapattheThresholdof

BlakHoleFormationI&II,"inProeedingsofthe8th

MarelGrossmannMeetingonGR,Ed.T.Piran,22-27,

June, 1997, Jerusalem, Israel(WorldSienti, 1997),

pp.723-727.

[20℄ E.W. Hirshmannand D. M. Eardley, Phys.Rev.D

51,4198(1995);ibid.D52,5850(1995);ibid.D56,4696

(1997); D. M. Eardley, E. W. Hirshmann, and J. H.

Horne, ibid., D52, 5397 (1995); R. S. Hamade and J.

M. Stewart, Class. Quantum Grav. 13, 497 (1996); S.

Liebling,Phys.Rev.D58,084015 (1998).

[21℄ S.L. Liebling and M. W.Choptuik, Phys.Rev.Lett.

77,1424 (1996); S.L.Liebling, M.A. Thesis, The

Uni-versityofTexasatAustin,1995(unpublished).

[22℄ M.W.Choptuik,T.Chmajand P.Bizon,Phys.Rev.

Lett. 77, 424 (1996); C. Gundlah, Phys. Rev. D55,

6002 (1997); M. W. Choptuik, E. W. Hirshmann, R.

(9)

[23℄ C.GundlahandJ.M.Martn-Gara,Phys.Rev.D54,

7353-7360(1996);S.HodandT.Piran,ibid.,D55,3485

(1997).

[24℄ P.R.Brady,C.M.Chambers,S.M.C.V.Gonalves,

Phys.Rev.D56,6057(1997).

[25℄ J.C. Niemeyer and K. Jedamzik, Phys. Rev. Lett.

80, 5481 (1997); Phys. Rev. D59, 124013 (1999); J.

Yokoyama,ibid.,D58,083510(1998);K.Jedamzikand

J.C. Niemeyer, ibid., D59, 124014 (1999); A.M.Green

andA.R.Liddle,ibid.,D60,063509 (1999).

[26℄ P. Bizon and T. Chmaj,

Phys.Rev.D58,041501(1998),AtaPhys.Polon.B29,

1071 (1998),gr-q/9802002; \Remarkon formationof

olored blakholesviane tuning,"gr-q/9906070; P.

Bizon,T.Chmaj,andZ.Tabor,Phys.Rev.D59,104003

(1999).

[27℄ G.Rein, A. D. Rendall, and J.Shaeer, Phys. Rev.

D58,044007(1998).

[28℄ I.Olabarrieta,\CritialCollapseofCollisionless

mat-terinSpherialSymmetry,"gr-q/0012059.

[29℄ M. Christensen, V.P. Frolov, and A.L.Larsen, Phys.

Rev.D58,085005(1998);V.P.Frolov,A.L.Larsen,and

M.Christensen,ibid.,D59,125008(1999).

[30℄ D.Garnkle,C.Cutler,andG.C.Dunan,Phys.Rev.

D60,104007(1999).

[31℄ S.L.Liebling,Phys.Rev.D60,061502(1999);D.

Mai-sonandS.L.Liebling,Phys.Rev.Lett.83,5218(1999);

P. Bizon, \Equivariant Self-Similar wave Maps from

MinkowskiSpaetimeinto3-Sphere,"math-ph/9910026;

P. Bizon, T. Chmaj, and Z. Tabor, \Dispersion and

Collapse of Wave Maps," math-ph/9912009; P. Bizon

and A. Wasserman, \Self-Similar Spherially

Symmet-ri Wave Maps Coupled to Gravity," gr-q/0006034;

S.L. Liebling, E.W. Hirshmann, and J. Isenberg,

\Critial Phenomena in Non-linear Sigma Models,"

gr-q/9911020; S. Husa, C. Lehner, M. Purrer, J.

Thornburg, and P.C. Aihelburg, \Type II Critial

Collapse of a Self-Gravitating Non-linear -Model,"

gr-q/0002067.

[32℄ A.Z. Wang, \Critial Collapse in Salar-Tensor and

Non-linear Gravity Theories: A Universal Class,"

gr-q/9901044.

[33℄ S.H.HawleyandandM.W.Choptuik,Phys.Rev.D62,

104024(2000).

[34℄ Y.PelegandA.R.Steif,Phys.Rev.D51,R3992(1995);

D. Birminghan and S. Sen, Phys.Rev. Lett. 84, 1074

(2000); F.Pretorius and M. W.Choptuik, Phys. Rev.

D62, 104012 (2000); L.M. Burko, ibid., D62, 127503

(2000); D. Garnkle, D63, 044007 (2001); V. Husain

andM.Olivier,Class.QuantumGrav.18,L1(2001);G.

Clementand A.Fabbri,\Analytial Treatment of

Crit-ial Collapse in 2+1 Dimensional AdS Spaetimes,"

gr-q/0101073; D.Birminghan,\Choptuik Salingand

QuasinormalModesintheAdS=CFTCorrespondene,"

hep-th/0101194.

[35℄ Thedenitionofdisreteself-similaritywasrstgiven

by Gundlah in C. Gundlah Phys. Rev. D55, 695

[36℄ Solutions of the homotheti (or the rst kind)

self-similaritywererststudiedbyCahillandTaubin

Gen-eralRelativityforaperfetuid[A.H.CahillandM.E.

Taub, Commu. Math. Phys. 21, 1 (1971).℄, and later

Carter and Henriksen [B. Carter and R.N. Henriksen,

Ann. Physique Suppl. 14, 47 (1989)℄ generalized this

asetoamoregeneralone,theso-alledkinemati

self-similarity,whihinludestheself-similaritiesoftherst,

seond, and zeroth kinds. For details of the latter, we

refer readerstoA.A. Coley,Class.QuantumGrav. 14,

87(1997);P.M.BenoitandA.A.Coley,Class.Quantum

Grav.15,2397(1998),C.F.C.Brandt,L.-M.Lin,J.F.V.

da Roha, and A.Z. Wang, \Gravitational ollapse of

spherially symmetri perfet uid with kinemati

self-similarity,"preprint,gr-q/0105019(2001);toappearin

Inter.J.Mod.Phys.D.

[37℄ N. Goldenfeld, Letures on Phase Transitions and

theRenormalizationGroup(AddisonWesleyPublishing

Company, New York, 1992); J. M. Yeomans,

Statisti-alMehanisofPhase Transitions(OxfordUniversity

Press1992).

[38℄ J.D. Bekenstein, \Blak Holes: Classial

Proper-ties, Thermodynamis, and Heuristi Quantization,"

gr-q/9808028.

[39℄ A.Z.WangandH.P.deOliveira,Phys.Rev.D56,753

(1997).

[40℄ M.D.Roberts,Gen.Relativ.Grav.21,907(1989).

[41℄ P.R. Brady, Class. Quantum Grav. 11, 1255 (1994);

Phys.Rev.D51,4168(1995);Y.Oshiro,K.Nakamura,

andA.Tomimatsu,Prog.Theor.Phys.91,1265(1994);

L.M.Burko,Gen.Rel.Grav.29,259(1997);S.A.

Hay-ward,Class.QuantumGrav.17,4021(2000).

[42℄ A. V. Frolov, Phys.Rev. D56, 6433 (1997); ibid., D

59,104011(1999);ibid.,D61,084006(2000).

[43℄ E.PoissonandW.Israel,Phys.Rev.D41,1796(1990).

[44℄ W.Israel, NuovoCimento,B44, 1(1966); ibid.,B48,

463(E)(1967).

[45℄ P.C. Vaidya, Pro. Ind. Aad. Sienes, A33, 264

(1951).

[46℄ C. Barrabes and W. Israel, Phys. Rev. D43, 1129

(1991).

[47℄ M. Alubierre, G. Allen, B. Bruegmann, G.

Lanfer-mann,E.Seidel,W.M.Suen,andM.Tobias,Phys.Rev.

D61,041501(2000).

[48℄ C.Gundlah,Phys.Rev.D577075(1998);ibid.,D57,

7080 (1998); J. M. Martn-Gara and C. Gundlah,

ibid., D59, 064031 (1999); D. Garnkle, C. Gundlah,

andJ.M.Martn-Gara, ibid.,D59,104012(1999).

[49℄ Privateommuniations.

[50℄ T. Chiba and M. Siino, Mod. Phys. Lett. A12, 709

(1997).

[51℄ S.AyalandT.Piran,Phys.Rev.D56,4768 (1997).

(10)

[53℄ P.C. Argyres, inContributionsfrom the G1 Working

Group at theAPS Summer Study on Partileand

Nu-lear Astrophysis and Cosmology in the Next

Millen-nium,Snowmass,Colorado,June29-July14, 1994,p.

23,astro-ph/9412046.

[54℄ D. Garnkle, Phys. Rev. D56, R3169 (1997); D.

Garnkleand C.Gundlah,Class. QuantumGrav.16,

4111 (1999); D. Garnkle and K. Meyer, Phys. Rev.

D59,064003(1999).

[55℄ R.H.PrieandJ.Pullin,Phys.Rev.D54,3792(1996).

[56℄ B.K. Harrison, K.P. Thorne, M. Wakano, and J.A.

Wheeler,GravitationTheoryandGravitationalCollapse

Imagem

Figure 1. The initial onguration of the massless salar
Figure 2. The phase spae of the dynami system of the
Figure 4. The mass M(r) of ollapsing spherial shell.
Figure 5. The orresponding Penrose diagrams. (a) The

Referências

Documentos relacionados

the generalized GDH sum rule, whih are of interest to issues related to the spin of the proton as.. measured at high energies,

model: the presene of nonbaryoni \dark matter&#34; in galaxies.. Using Newton's law

We onsider the saling behavior in the ritial domain of superondutors at zero external

trap and laser system being built in Rio, to trap and perform high resolution spetrosopy on.. old hydrogen, is presented along with a disussion on the tehniques and

An ultrastati spaetime admits a globally dened oordinate system in whih the omponents of the metri tensor are time-. independent and the onditions g00 = 1 and goi =

Four-jet Higgs boson andidate with a reonstruted Higgs boson mass of 112.9 GeV/ 2. All four jets

and magneti harges, being dierent from monopoles (magneti or eletri), the topologial ontent.. of monopoles, surfae terms, Dira strings and hiral struture in the inlusion of

the ultra high energy osmi rays [21℄. From