-
改写二阶为一阶:
令 k=∂tϕ+g0ig00∂iϕ,ξi=∂iϕ, 那么 ∂tξi=∂i∂tϕ=∂i(k−g0ig00∂iϕ)
记 T=2Nφcd(gijΦicaΦjdb−ΠcaΠbd−φefΓaceΓbdf)−12NtctdΠcdΠab−NtcΠcigijΦjab−2N∇(aHb)
令 Γb=Hb 就是广义调和,令 Φiab=∂iφab, −NΠab=∂tφab−NiΦiab , t=(t0,t1,t2,t3)=(1N,−N1N,−N2N,−N3N)
现在的变量有 {φab,Πab,Φiab,ϕ,k,ξi}
∂tφab−Nk∂kφab∂tΠab−Nk∂kΠab+Ngki∂kΦiab∂tΦiab−Nk∂kΦiab∂tϕ−Ni∂iϕ∂tk−Ni∂ik−gijN2∂iξj∂tξi−Nj∂jξi−∂ik=−NΠab=T+(∂aϕ)(∂bϕ)+φabV=12NtctdΦicdΠab+NgjktcΦijkΦkab=k=Ngij[N(Hj+gklΓikl)−∂jN]+∂V∂ϕ=∂i(Nj)ξj
边界条件:我们先选择度量:
ds2=−N2dt2+γ(dr+βdt)2=(−N2+β2γ)dt2+γdr2+2βγdrdt
ϕ(0,r)N(0,r)β(0,r)γ′(0,r)|r=0γ(0,r)|r→∞∂rφab∂rΠab=ϕ0e−(r−r0)2/Δ2=γ(0,r)−2=0=0=1=0=0
这个边值条件和施瓦希差多远?
⎛⎝⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜−N1000000000000−N1000000000000−N1000000000000−N1000000000000−N1000000000000−N1000000000g11N00−N1000000000g11N00−N1000000000g11N00−N1000000000000−N1000000000000−N1−10000000000−g11N2−N1⎞⎠⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟
特征值: {−N1(10),−N1−g11−−−√N,−N1+g11−−−√N}
特征向量:
{0,0,0,0,0,0,0,0,0,1,0,0},{0,0,0,0,0,1,0,0,0,0,0,0}, {0,0,0,0,1,0,0,0,0,0,0,0},{0,0,0,1,0,0,0,0,0,0,0,0}, {0,0,1,0,0,0,0,0,0,0,0,0},{0,1,0,0,0,0,0,0,0,0,0,0}, {1,0,0,0,0,0,0,0,0,0,0,0},{0,0,0,0,0,0,0,0,0,0,0,0}, {0,0,0,0,0,0,0,0,0,0,0,0},{0,0,0,0,0,0,0,0,0,0,0,0}, {0,0,0,0,0,0,0,0,0,0,g11−−−√N,1},{0,0,0,0,0,0,0,0,0,0,−g11−−−√N,1}
apparent horizon
θ=r∂tS+(1+r∂rS)(−gtrgrr+g2trg2rr−gttgrr−−−−−−−−√)=0
在本文简单情况下θ=−gtrgrr+g2trg2rr−gttgrr−−−−−−−−√=0
不把
ϕ
考虑进方程组,前9个未知变量
⎛⎝⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜⎜−N1000000000−N1000000000−N1000000000−N1000000000−N1000000000−N1000000g11N00−N1000000g11N00−N1000000g11N00−N1⎞⎠⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟⎟
特征向量是
{0,0,0,0,0,1,0,0,0}
,
{0,0,0,0,1,0,0,0,0}
,
{0,0,0,1,0,0,0,0,0}
,
{0,0,1,0,0,0,0,0,0}
,
{0,1,0,0,0,0,0,0,0}
,
{1,0,0,0,0,0,0,0,0}
,
{0,0,0,0,0,0,0,0,0}
,
{0,0,0,0,0,0,0,0,0}
,
{0,0,0,0,0,0,0,0,0}
特征值:
{−1,−1,−1,1,1,1,0,0,0}
- 另外一种方案:不把标量场
ϕ
放到一阶方程组里。
∂tφab−Nk∂kφab∂tΠab−Nk∂kΠab+Ngki∂kΦiab∂tΦiab−Nk∂kΦiab□ϕ=−NΠab=T+(∂aϕ)(∂bϕ)+φabV=12NtctdΦicdΠab+NgjktcΦijkΦkab=∂V∂ϕ
给出边界条件:
2√[k(cφd)a−12kaφcd]d⊥u1^−cdd⊥u0^abd⊥u2^kab(PcaPdb−12PabPcd)(d⊥u1^±cd)=0=0=0=0
- Lee的笔记:
Ek=Πk+Φkt=12(u1+k+u1−k)+12nk(u1+t−u1−t)+u2^kt
则nkEk=12(nku1+k+nku1−k+u1+t−u1−t)+nku2^kt(35)
关于 nku2^kt 这一项,注意: nkPikΦit=nk(δik−nink)Φit=0 ,所以这项必须消失。
而 Φij=12ni(u1+j−u1−j)+u2^ij,Bk=ϵijkΦij=12ϵijkni(u1+j−u1−j)+ϵijku2^ij
那么 nkBk=
12ϵijknkni(u1+j−u1−j)+nkϵijku2^ij=12ϵijknkni(u1+j−u1−j)+ϵijknku2^ij(36)
关于 12ϵijknkni(u1+j−u1−j) 这项,注意: (A×B)i=ϵijkAjBk=ϵijkAjBk ,所以这项必须为0.以下推导(37)式
PkiEk=12Pki(u1+k+u1−k)+Pkiu2^kt=12Pki(u1+k+u1−k)+u2^it
ϵjkinjBk=ϵjkinjϵabku2^ab+12ϵjkinjϵabkna(u1+b−u1−b)
ϵjkinjBk=ϵjkinjϵabku2^ab+12ϵjkinjϵabkna(u1+b−u1−b)=ϵjikϵabknju2^ab+12ϵijkϵabknjna(u1+b−u1−b)=nju2^ab(δajδbi−δbjδai)+12njna(δaiδbj−δbiδaj)=−nju2^ij−12Pji(u1+j)+12Pji(u1−j)
可得:PkiEk+ϵjkinjBk=Pkiu1−k+u2^it+nku2^ik(37)