/*-*MACSYMA-*-*/ /* * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License as * published by the Free Software Foundation; either version 2 of * the License, or (at your option) any later version. * * This program is distributed in the hope that it will be * useful, but WITHOUT ANY WARRANTY; without even the implied * warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR * PURPOSE. See the GNU General Public License for more details. * * Comments: * * The ctensor package was downcased, cleaned up, and moving frames * functionality was added by Viktor Toth (http://www.vttoth.com/). * * As of November, 2004, the naming conventions in this package now * correspond with the naming conventions in commercial MACSYMA. */ if get('ctensor,'version)#false then error("CTENSOR already loaded!")$ /*********************************** PACKAGE INITIALIZATION ***********************************/ tensorkill:true$ define_variable(dim,4,fixnum,"The dimension of the problem.")$ define_variable(diagmetric,false,boolean, "True if the metric entered is a diagonal matrix.")$ define_variable(ctrgsimp,false,boolean, "Apply trigonometric simplifications (TRIGSIMP).")$ define_variable(cframe_flag,false,boolean, "Perform computations in a moving frame.")$ define_variable(ctorsion_flag,false,boolean,"Apply torsion")$ define_variable(cnonmet_flag,false,boolean,"Apply nonmetricity")$ define_variable(derivabbrev,true,boolean)$ define_variable(ctayswitch,false,boolean)$ define_variable(ratchristof,true,boolean)$ define_variable(rateinstein,true,boolean)$ define_variable(ratriemann,true,boolean)$ define_variable(ratweyl,true,boolean)$ init_ctensor():= ( tensorkill:true, dim:4, diagmetric:false, setflags(), kill(ct_coords,lg,ug,lcs,mcs,geod,ric,uric,tracer,ein,lein, riem,lriem,uriem,kinvariant,lfg,ufg,fr,fri,fb,tr,kt,nm,nmc), lfg:diagmatrix(dim,1), lfg[1,1]:-1, done )$ _ug():=if cframe_flag then 'ufg else 'ug; _lg():=if cframe_flag then 'lfg else 'lg; /*********************************** HELPER FUNCTIONS ***********************************/ setflags():=(derivabbrev:true,ctayswitch:false, ratchristof:true,rateinstein:true,ratriemann:true,ratweyl:true)$ readvalue(message,pred,[badboy]):= block([value], loop, value:read(message), if apply(pred,[value])=true then return(value), if badboy#[] then apply('print,badboy), go(loop))$ modedeclare(function(yesp),boolean)$ yesp([messages]):=block ( [reply], loop, reply:apply('read,messages), if member(reply,['yes,'ye,'y,'yes,'yes,'y]) then return(true), if member(reply,['no,'n,'no,'no,'n]) then return(false), print("Please reply YES or NO."), go(loop) )$ resimp(exp):=apply('ev,[exp,'noeval,'simp])$ kdelt(i,j):=(modedeclare([i,j],fixnum),if i = j then 1 else 0)$ /* Optional Taylor-series approximation */ ctaylor(x):= if ctayswitch then ratdisrep(taylor(x,ctayvar,ctaypt,ctaypov-1)) else x$ /* This function transforms second-rank covariant tensors given the components and the transformation law, which is in the form Yi=Yi(X1,...,Xdim) */ ctransform(qxyz):=block ( [ttemp,omtemp,vartemp,newtemp], for i thru dim do for j thru dim do ( ttemp[i,j]:qxyz[i,j], for k thru dim do ttemp[i,j]:subst(omtemp[k],ct_coords[k],ttemp[i,j]) ), for i thru dim do vartemp[i]:read("Transform #",i), for i thru dim do for j thru dim do for k thru dim do ttemp[i,j]:subst(vartemp[k],omtemp[k],ttemp[i,j]), for a thru dim do for b thru dim do newtemp[a,b]:txyzsum(vartemp,ct_coords,a,b,ttemp), genmatrix(newtemp,dim,dim) )$ txyzsum(vartemp,ct_coords,a,b,ttemp):=block( [temp:0], for i from 1 thru dim do for j from 1 thru dim do temp:temp+diff(vartemp[i],ct_coords[a])*diff(vartemp[j],ct_coords[b])* ttemp[i,j], temp )$ /* Check diagonality of a matrix or (2D) array */ diagmatrixp(mat,nlim):= ( modedeclare(nlim,fixnum), block ( [diagflag:true], for i thru nlim do for j thru nlim do if i#j and mat[i,j]#0 then return(diagflag:false), diagflag ) )$ /* Check symmetry of a matrix or (2D) array */ symmetricp(m,n):= ( modedeclare(n,fixnum), block ( [symflag:true], if n#1 then for i thru n-1 do for j from i+1 thru n do if m[j,i]#m[i,j] then return(symflag:false), symflag ) )$ /* Count the number of terms in a (2D) array */ ntermst(f):=for i thru dim do for j thru dim do print([[i,j],nterms(f[i,j])])$ /* Display the contents of a tensor */ cdisplay(_t,[_k]):=block ( [_len,_tmp,_i,_j,_n], if member(_t,arrays) then ( _tmp : funmake(arrayinfo, [_t]), _len : ev(_tmp)[2] - length(_k), if _len > 2 then ( for _n thru dim do apply(cdisplay,append([_t],append(_k,[_n]))) ) else if _len = 2 then block ( [_tmp:zeromatrix(dim,dim)], for _i thru dim do for _j thru dim do _tmp[_i,_j]:arrayapply(_t,append(_k,[_i,_j])), disp((if length(_k) > 0 then arrayapply(nounify(_t),_k) else nounify(_t))=_tmp) ) else if _len = 1 then (true) else print("Array is empty") ) else disp(_t=ev(_t)), done )$ /* Delete an element */ deleten(l,n):= ( modedeclare(n,fixnum), block ( [len], modedeclare(len,fixnum), if l = [] then l else ( len:length(l), if n>len or n<1 then error("Second argument out of range") else if n=1 then rest(l) else if n=len then rest(l,-1) else append(rest(l,n-len-1),rest(l,n)) ) ) )$ /*********************************** SETTING UP THE METRIC ***********************************/ csetup():=block ( [], if not tensorkill then error("Type KILL(ALL)\; and then TENSORKILL:TRUE\; before you enter a new metric."), tensorkill:false, setflags(), dim:mode_identity(fixnum, readvalue("Enter the dimension of the coordinate system:", lambda([v], if integerp(v) then block ( [u:v], modedeclare(u,fixnum), if u>0 then true ) ), "Must be a positive integer!") ), ct_coords:if dim=2 then [x,y] else if dim = 3 then [x,y,z] else if dim = 4 then [x,y,z,t] else read("Enter a list containing the names of the coordinates in order"), if yesp("Do you wish to change the coordinate names?") then ct_coords: read("Enter a list containing the names of the coordinates in order"), if length(ct_coords)#dim then error("Length of the coordinate list is not equal to the dimension"), readvalue("Do you want to 1. Enter a new metric? 2. Enter a metric from a file? 3. Approximate a metric with a Taylor series?", lambda([opt], if opt = 1 then(newmet(),true) else if opt = 2 then(filemet(),true) else if opt = 3 then(sermet(),true) else false ), "Invalid option, please enter 1, 2, or 3." ), done )$ /* Enter a new metric by hand */ newmet():=block ( [], lg:entermatrix(dim,dim), read("Enter functional dependencies with DEPENDS or 'N' if none"), if yesp("Do you wish to see the metric?") then disp(lg), cmetric() )$ /* Read a metric from a file */ filemet():=block ( [file,fpos], file:read("Specify the file as [filename1,filename2,directory]?"), fpos:(print("What is the ordinal position of the metric in this file?"), read("(Note, the name LG must be assigned to your metric in the file)")), apply(batch,[file,off,fpos,fpos]),cmetric() )$ /* Approximate the metric with a Taylor series */ sermet():=block ( [], ctayswitch:true, ctayvar:read("Enter expansion parameter"), ctaypov:read("Enter minimum power to drop"), ctaypt:read("Enter the point to expand the series around"), if yesp("Is the metric in a file?") then filemet() else newmet() )$ /* Computing the frame field from the inverse frame field (user-defined) */ tmetric(disp):=block( [], /* fr are the contravariant frame vector components. fri are covariant. i.e., i fr[a,i] = e , fri[a,i] = e (a) (a)i */ fr:lfg.(transpose(fri)^^(-1)), ufg:lfg^^(-1), lg:transpose(lfg.fri).fri, if ctrgsimp then lg:trigsimp(lg) else lg:ratsimp(lg), if ratfac then lg:factor(lg), if disp then ldisplay(lg) ); /* Computing the contravariant metric or the frame field */ cmetric([dis]):=if cframe_flag then tmetric(if dis#[] then dis[1] else false) else block ( [], if length(lg)#dim or length(transpose(lg))#dim then error("The rank of the metric is not equal to the dimension of the space"), /* else if not symmetricp(lg,dim) then error("The metric tensor must be symmetric!"),*/ diagmetric:diagmatrixp(lg,dim), gdet:factor(determinant(lg)), ug:ident(length(first(lg))), if diagmetric then for ii:1 thru length(first(lg)) do ug[ii,ii]:1/lg[ii,ii] else ug:block([detout:true],transpose(lg^^(-1))), if ratfac then ug:ratsimp(ug), if dis#[] and dis[1]#false then if yesp("Do you wish to see the metric inverse?") then ldisp(ug), ug:resimp(ug), done )$ ct_coordsys(coordinate_system, [extra_args]):= ( if coordinate_system='cartesian2d then ( dim:2, ct_coords:[x,y], if cframe_flag then ( fri:ident(2), lfg:ident(2) ) else lg:ident(2) ) else if coordinate_system='polar then ( dim:2, ct_coords:[r,phi], assume(r>=0), if cframe_flag then ( fri:matrix([cos(phi),-sin(phi)*r], [sin(phi), cos(phi)*r]), lfg:ident(2) ) else lg:matrix([1,0],[0,r^2]) ) else if coordinate_system='elliptic then ( dim:2, ct_coords:[u,v], declare(e,constant), if cframe_flag then ( fri:matrix([e*sinh(u)*cos(v),-e*cosh(u)*sin(v)], [e*cosh(u)*sin(v), e*sinh(u)*cos(v)]), lfg:ident(2) ) else lg:matrix([e^2*(cosh(u)^2-cos(v)^2), 0], [ 0, e^2*(cosh(u)^2-cos(v)^2)]) ) else if coordinate_system='confocalelliptic then block( [tmp:sqrt((u^2-1)*(1-v^2))], dim:2, ct_coords:[u,v], declare(e,constant), if cframe_flag then ( fri:matrix([ e*v, e*u], [e*u*(1-v^2)/tmp, e*v*(1-u^2)/tmp]), lfg:ident(2) ) else lg:matrix([e^2*(u^2-v^2)/(u^2-1), 0], [0, e^2*(v^2-u^2)/(v^2-1)]) ) else if coordinate_system='bipolar then block( [tmp:(cosh(v)-cos(u))^2], dim:2, ct_coords:[u,v], declare(e,constant), if cframe_flag then ( fri:matrix([ -e*sin(u)*sinh(v)/tmp, e*(1-cos(u)*cosh(v))/tmp], [e*(cos(u)*cosh(v)-1)/tmp, -e*sin(u)*sinh(v)/tmp]), lfg:ident(2) ) else lg:matrix([e^2/tmp,0],[0,e^2/tmp]) ) else if coordinate_system='parabolic then ( dim:2, ct_coords:[u,v], if cframe_flag then ( fri:matrix([u,-v],[v,u]), lfg:ident(2) ) else lg:matrix([u^2+v^2, 0], [0, u^2+v^2]) ) else if coordinate_system='cartesian3d then ( dim:3, ct_coords:[x,y,z], if cframe_flag then ( fri:ident(3), lfg:ident(3) ) else lg:ident(3) ) else if coordinate_system='polarcylindrical then ( dim:3, ct_coords:[r,theta,z], if cframe_flag then ( fri:matrix([cos(theta), -r*sin(theta), 0], [sin(theta), r*cos(theta), 0], [ 0, 0, 1]), lfg:ident(3) ) else lg:matrix([1, 0, 0], [0, r^2, 0], [0, 0, 1]) ) else if coordinate_system='ellipticcylindrical then ( dim:3, ct_coords:[u,v,z], if cframe_flag then ( fri:matrix([e*sinh(u)*cos(v), -e*cosh(u)*sin(v), 0], [e*cosh(u)*sin(v), e*sinh(u)*cos(v), 0], [ 0, 0, 1]), lfg:ident(3) ) else lg:matrix([e^2*(sin(v)^2+sinh(u)^2), 0, 0], [ 0, e^2*(sin(v)^2+sinh(u)^2), 0], [ 0, 0, 1]) ) else if coordinate_system='confocalellipsoidal then ( dim:3, ct_coords:[u,v,w], declare([e,f,g],constant), if cframe_flag then ( fri:matrix([-sqrt((v-e^2)*(w-e^2)/(f^2-e^2)/(g^2-e^2)/(e^2-u))/2, -sqrt((u-e^2)*(w-e^2)/(f^2-e^2)/(g^2-e^2)/(e^2-v))/2, -sqrt((u-e^2)*(v-e^2)/(f^2-e^2)/(g^2-e^2)/(e^2-w))/2], [ sqrt((v-f^2)*(w-f^2)/(f^2-e^2)/(g^2-f^2)/(u-f^2))/2, sqrt((u-f^2)*(w-f^2)/(f^2-e^2)/(g^2-f^2)/(v-f^2))/2, sqrt((u-f^2)*(v-f^2)/(f^2-e^2)/(g^2-f^2)/(w-f^2))/2], [-sqrt((v-g^2)*(w-g^2)/(g^2-e^2)/(g^2-f^2)/(g^2-u))/2, -sqrt((u-g^2)*(w-g^2)/(g^2-e^2)/(g^2-f^2)/(g^2-v))/2, -sqrt((u-g^2)*(v-g^2)/(g^2-e^2)/(g^2-f^2)/(g^2-w))/2]), lfg:ident(3) ) else lg:matrix([(v-u)*(w-u)/4/(e^2-u)/(u-f^2)/(u-g^2), 0, 0], [0, (v-u)*(w-v)/4/(v-e^2)/(v-f^2)/(v-g^2), 0], [0, 0, (w-u)*(w-v)/4/(e^2-w)/(w-f^2)/(w-g^2)]) ) else if coordinate_system='bipolarcylindrical then block( [tmp:(cosh(v)-cos(u))^2], dim:3, ct_coords:[u,v,z], declare(e,constant), if cframe_flag then ( fri:matrix([e*sin(u)*sinh(v)/tmp, e*(1-cos(u)*cosh(v))/tmp, 0], [e*(cos(u)*cosh(v)-1)/tmp, -e*sin(u)*sinh(v)/tmp, 0], [ 0, 0, e/tmp]), lfg:ident(3) ) else lg:matrix([e^2/tmp, 2*e^2*sin(u)*sinh(v)*(1-cos(u)*cosh(v))/tmp, 0], [2*e^2*sin(u)*sinh(v)*(1-cos(u)*cosh(v))/tmp, e^2/tmp, 0], [ 0, 0, e^2/tmp^2]) ) else if coordinate_system='paraboliccylindrical then ( dim:3, ct_coords:[u,v,z], if cframe_flag then ( fri:matrix([u,-v,0],[v,u,0],[0,0,1]), lfg:ident(3) ) else lg:matrix([u^2+v^2, 0, 0],[0, u^2+v^2, 0],[0, 0, 1]) ) else if coordinate_system='paraboloidal then ( dim:3, ct_coords:[u,v,phi], if cframe_flag then ( fri:matrix([-sin(phi)*u*v, cos(phi)*v, cos(phi)*u], [ cos(phi)*u*v, sin(phi)*v, sin(phi)*u], [ 0, u, -v]), lfg:ident(3) ) else lg:matrix([u^2*v^2, 0, 0],[0, u^2+v^2, 0], [0, 0, u^2+v^2]) ) else if coordinate_system='conical then ( dim:3, ct_coords:[u,v,w], declare([e,f],constant), if cframe_flag then ( fri:matrix([v*w/e/f, u*w/e/f, u*v/e/f], [-u*w/sqrt((f^2-e^2)*(u^2-e^2)/(e^2-v^2))/e, -v*w/sqrt((f^2-e^2)*(v^2-e^2)/(e^2-u^2))/e, sqrt((u^2-e^2)*(v^2-e^2)/(e^2-f^2))/e], [ u*w/sqrt((f^2-e^2)*(u^2-f^2)/(v^2-f^2))/f, v*w/sqrt((f^2-e^2)*(v^2-f^2)/(u^2-f^2))/f, sqrt((u^2-f^2)*(v^2-f^2)/(f^2-e^2))/f]), lfg:ident(3) ) else lg:matrix([(v-u)*(v+u)*w^2/((u-e)*(u+e)*(u-f)*(u+f)), 0, 0], [0, (u-v)*(u+v)*w^2/((v-e)*(v+e)*(v-f)*(v+f)), 0], [0, 0, 1]) ) else if coordinate_system='toroidal then ( dim:3, ct_coords:[u,v,phi], declare(e,constant), if cframe_flag then ( fri:matrix([-e*sin(phi)*sinh(v)/(cosh(v)-cos(u)), -e*cos(phi)*sin(u)*sinh(v)/(cosh(v)-cos(u))^2, -e*cos(phi)*(cos(u)*cosh(v)-1)/(cosh(v)-cos(u))^2], [ e*cos(phi)*sinh(v)/(cosh(v)-cos(u)), -e*sin(phi)*sin(u)*sinh(v)/(cosh(v)-cos(u))^2, -e*sin(phi)*(cos(u)*cosh(v)-1)/(cosh(v)-cos(u))^2], [ 0, e*(cos(u)*cosh(v)-1)/(cosh(v)-cos(u))^2, -e*sin(u)*sinh(v)/(cosh(v)-cos(u))^2]), lfg:ident(3) ) else lg:matrix([e^2*sinh(v)^2/(cosh(v)-cos(u))^2, 0, 0], [0, e^2/(cosh(v)-cos(u))^2, 0], [0, 0, e^2/(cosh(v)-cos(u))^2]) ) else if coordinate_system='spherical then ( dim:3, ct_coords:[r,theta,phi], assume(r>=0), if cframe_flag then ( fri:matrix([1,0,0],[0,r,0],[0,0,r*sin(theta)]), lfg:ident(3) ) else lg:matrix([1,0,0],[0,r^2,0],[0,0,r^2*sin(theta)^2]) ) else if coordinate_system='oblatespheroidal then ( dim:3, ct_coords:[u,v,phi], declare(e,constant), if cframe_flag then ( fri:matrix([abs(e)*sqrt(sin(v)^2+sinh(u)^2), 0, 0], [0, abs(e)*sqrt(sin(v)^2+sinh(u)^2), 0], [0, 0, abs(e)*cosh(u)*abs(cos(v))]), lfg:ident(3) ) else lg:matrix([e^2*(sin(v)^2+sinh(u)^2), 0, 0], [0, e^2*(sin(v)^2+sinh(u)^2), 0], [0, 0, e^2*cosh(u)^2*cos(v)^2]) ) else if coordinate_system='oblatespheroidalsqrt then ( dim:3, ct_coords:[u,v,phi], declare(e,constant), if cframe_flag then ( fri:matrix([abs(e)*sqrt((u^2-v^2)/(u^2-1)), 0, 0], [0, abs(e)*sqrt((u^2-v^2)/(u^2-1)), 0], [0, 0, abs(e*u*v)]), lfg:ident(3) ) else lg:matrix([e^2*(u^2-v^2)/(u^2-1), 0, 0], [0, e^2*(u^2-v^2)/(u^2-1), 0], [0, 0, e^2*u^2*v^2]) ) else if coordinate_system='prolatespheroidal then ( dim:3, ct_coords:[u,v,phi], declare(e,constant), if cframe_flag then ( fri:matrix([abs(e)*sqrt(sin(v)^2+sinh(u)^2), 0, 0], [0, abs(e)*sqrt(sin(v)^2+sinh(u)^2), 0], [ 0, 0, abs(e*sinh(u)*sin(v))]), lfg:ident(3) ) else lg:matrix([e^2*(sin(v)^2+sinh(u)^2), 0, 0], [0, e^2*(sin(v)^2+sinh(u)^2), 0], [0, 0, e^2*sin(v)^2*sinh(u)^2]) ) else if coordinate_system='prolatespheroidalsqrt then ( dim:3, ct_coords:[u,v,phi], declare(e,constant), if cframe_flag then ( fri:matrix([abs(e)*sqrt((v^2-u^2)/(1-u^2)), 0, 0], [0, abs(e)*sqrt((v^2-u^2)/(v^2-1)), 0], [0, 0, abs(e)*sqrt((1-u^2)*(v^2-1))]), lfg:ident(3) ) else lg:matrix([e^2*(v^2-u^2)/(1-u^2), 0, 0], [0, e^2*(v^2-u^2)/(v^2-1), 0], [0, 0, e^2*(1-u^2)*(v^2-1)]) ) else if coordinate_system='ellipsoidal then ( dim:3, ct_coords:[r,theta,phi], declare([a,b,c],constant), if cframe_flag then error("No frame fields yet") else lg:matrix( [(a^2*cos(phi)^2+b^2*sin(phi)^2)*sin(theta)^2+c^2*cos(theta)^2, (a^2*cos(phi)^2+b^2*sin(phi)^2-c^2)*r*cos(theta)*sin(theta), (b^2-a^2)*cos(phi)*sin(phi)*r*sin(theta)^2], [(a^2*cos(phi)^2+b^2*sin(phi)^2-c^2)*r*cos(theta)*sin(theta), r^2*((a^2*cos(phi)^2+b^2*sin(phi)^2)*cos(theta)^2+c^2*sin(theta)^2), (b^2-a^2)*cos(phi)*sin(phi)*r^2*cos(theta)*sin(theta)], [(b^2-a^2)*cos(phi)*sin(phi)*r*sin(theta)^2, (b^2-a^2)*cos(phi)*sin(phi)*r^2*cos(theta)*sin(theta), (a^2*sin(phi)^2+b^2*cos(phi)^2)*r^2*sin(theta)^2]) ) else if coordinate_system='cartesian4d then ( dim:4, ct_coords:[x,y,z,t], if cframe_flag then fri:ident(4) else lg:ident(4), if cframe_flag then lfg:ident(4) ) else if coordinate_system='spherical4d then ( dim:4, ct_coords:[r,theta,eta,phi], /* declare(kurv,constant),*/ if cframe_flag then ( fri:matrix([1 /* sqrt(1/(1-kurv*r^2)) */, 0, 0, 0], [ 0, r, 0, 0], [ 0, 0, r*sin(theta), 0], [ 0, 0, 0, r*sin(eta)*sin(theta)]), lfg:ident(4) ) else lg:matrix([1 /* /(1-kurv*r^2) */, 0, 0, 0], [0, r^2, 0, 0], [0, 0, r^2*sin(theta)^2, 0], [0, 0, 0, r^2*sin(eta)^2*sin(theta)^2]) ) else if coordinate_system='exteriorschwarzschild then ( declare(m,constant), assume(r>2*m), dim:4, ct_coords:[t,r,theta,phi], if cframe_flag then ( fri:matrix( [sqrt((r-2*m)/r), 0,0, 0], [ 0,sqrt(r/(r-2*m)),0, 0], [ 0, 0,r, 0], [ 0, 0,0,sin(theta)*r]), lfg:ident(4), lfg[1,1]:-1 ) else lg:matrix( [(2*m-r)/r, 0, 0, 0], [ 0,r/(r-2*m), 0, 0], [ 0, 0,r^2, 0], [ 0, 0, 0,sin(theta)^2*r^2]) ) else if coordinate_system='interiorschwarzschild then ( dim:4, ct_coords:[t,z,u,v], declare([m],constant), assume(t<2*m), if cframe_flag then ( fri:matrix( [sqrt(t/(2*m-t)), 0, 0, 0], [ 0, sqrt((2*m-t)/t), 0, 0], [ 0, 0, t, 0], [ 0, 0, 0, sin(u)*t]), lfg:ident(4), lfg[1,1]:-1 ) else lg:matrix([-t/(2*m-t), 0, 0, 0], [ 0, (2*m-t)/t, 0, 0], [ 0, 0, t^2, 0], [ 0, 0, 0, sin(u)^2*t^2]) ) else if coordinate_system='kerr_newman then ( dim:4, declare([a,m,e],constant), rho:sqrt(r^2+(a*cos(theta))^2), delta:a^2-2*m*r+r^2+e^2, ct_coords:[ct,r,theta,phi], if cframe_flag then ( fri:matrix( [sqrt(delta)/rho, 0, 0, -a*sqrt(delta)*sin(theta)^2/rho], [0, rho/sqrt(delta), 0, 0], [0, 0, rho, 0], [-a*sin(theta)/rho,0, 0, (r^2+a^2)*sin(theta)/rho]), lfg:ident(4), lfg[1,1]:-1 ) else lg:matrix( [(a^2*sin(theta)^2-delta)/rho^2,0, 0, a*sin(theta)^2*(delta-r^2-a^2)/rho^2], [0, rho^2/delta, 0, 0], [0, 0,rho^2, 0], [a*sin(theta)^2*(delta-r^2-a^2)/rho^2,0,0, ((r^2+a^2)^2-a^2*delta*sin(theta)^2)*sin(theta)^2/rho^2]) ) /* Now some undocumented metrics from metrics.mac and other sources */ else if coordinate_system='standard then ( dim:4, remove([a,d],constant), depends([a,d],x), ct_coords:[x,y,z,t], if cframe_flag then error("No frames implementation") else lg:matrix([a,0,0,0],[0,x^2,0,0],[0,0,x^2*sin(y)^2,0],[0,0,0,d]) ) else if coordinate_system='nondiagonal3d then ( dim:3, ct_coords:[x,y,z], if cframe_flag then error("No frames implementation") else lg:lg:matrix([x^2,y,0],[y,y^2,0],[0,0,z]) ) else if coordinate_system='alternateflat4d then ( dim:4, ct_coords:[x,y,z,t], if cframe_flag then error("No frames implementation") else lg:matrix([1-t^2,0,0,1-t*x],[0,1,0,0],[0,0,1,0],[1-t*x,0,0,1-x^2]) ) else if coordinate_system='eddington_finkelstein then ( dim:4, ct_coords:[r,theta,phi,v], if cframe_flag then error("No frame fields yet") else lg:matrix([0, 0, 0, 1], [0, r^2, 0, 0], [0, 0, sin(theta)^2*r^2, 0], [1, 0, 0, (2*m-r)/r]) ) else if coordinate_system='isotropicschwarzschild then ( dim:4, ct_coords:[t,r,theta,phi], if cframe_flag then ( fri:matrix( [(1-m/2/r)/(1+m/2/r), 0, 0, 0], [0, (1+m/2/r)^2, 0, 0], [0, 0, (1+m/2/r)^2*r, 0], [0, 0, 0, (1+m/2/r)^2*r*sin(theta)]), lfg:ident(4), lfg[1,1]:-1 ) else lg:matrix([-((1-m/2/r)/(1+m/2/r))^2, 0, 0, 0], [ 0, (1+m/2/r)^4, 0, 0], [ 0, 0, (1+m/2/r)^4*r^2, 0], [ 0, 0, 0, (1+m/2/r)^4*r^2*sin(theta)^2]) ) else if coordinate_system='vacuum then ( dim:4, ct_coords:[x,y,z,t], if cframe_flag then error("No frames implementation") else lg:matrix([0,1,a*(2*y-t),1],[1,0,a*(2*x-t),1], [a*(2*y-t),a*(2*x-t),-6*a^2*(y+x)^2 +2*a^2*(2*x-t)*(2*y-t)-3/2,-a*(y+x+2*t)], [1,1,-a*(y+x+2*t),1/2]) ) else if coordinate_system='bondi then ( dim:4, ct_coords:[x,y,z,t], dependencies(a(x,y,t),b(x,y,t),c(x,y,t),d(x,y,t)), if cframe_flag then error("No frames implementation") else lg:matrix([0,0,0,%e^(2*b)],[0,-%e^(2*a)*x^2,0,%e^(2*a)*d*x^2], [0,0,-%e^(-2*a)*x^2*sin(y)^2,0], [%e^(2*b),%e^(2*a)*d*x^2,0,%e^(2*b)*c/x-%e^(2*a)*d^2*x^2]) ) else if coordinate_system='brans_dicke then ( dim:4, ct_coords:[x,y,z,t], if cframe_flag then error("No frames implementation") else lg:at(matrix([a,0,0,0],[0,a*x^2,0,0], [0,0,a*x^2*sin(y)^2,0],[0,0,0,-d]), [d=((1-b/x)/(1+b/x))^(2/l), a=(1+b/x)^4*((1-b/x)/(1+b/x))^(2*(l-c-1)/l), p=((1-b/x)/(1+b/x))^(c/l)]) /* ell:l=sqrt((c+1)^2-c*(1-w*c/2)) */ ) else if coordinate_system='lapides then ( /* RICCI FLAT METRIC FROM QUANTUM GRAVITY-FOUND BY ALAN LAPIDES- LOS ALAMOS APRIL 1982- PRESUME IT IS ONE OF THE HARRISON METRICS */ dim:4, ct_coords:[x,y,z,t], if cframe_flag then error("No frames implementation") else lg:at(matrix([1/c,0,0,0],[0,c*a,0,0],[0,0,c*b,0],[0,0,0,c*a]), [c=%e^-(2*atan(sinh(t))),a=(cosh(t)^2-sin(y)^2)^2/cosh(t)^2, b=cosh(t)^2*sin(y)^2]) ) else if coordinate_system='kantowski_sachs then ( dim:4, ct_coords:[r,theta,phi,t], if cframe_flag then error("No frames implementation") else ( lg:matrix([X(t)^2,0,0,0],[0,Y(t)^2,0,0], [0,0,Y(t)^2*sin(theta)^2,0],[0,0,0,-N^2]), if member('misner,extra_args) then lg:at(lg,[X(t)=exp(2*sqrt(3)*b(t)),Y(t)=exp(-2*sqrt(3)*o(t))])) ) else if listp(coordinate_system) then ( dim:length(coordinate_system)-1, ct_coords:coordinate_system[dim+1], if cframe_flag then ( lfg:ident(dim), fri:zeromatrix(dim,dim), for i thru dim do for j thru dim do fri[i,j]:diff(coordinate_system[i],ct_coords[j]) ) else ( lg:zeromatrix(dim,dim), for i thru dim do for j thru i do ( lg[i,j]:sum(diff(coordinate_system[k%],ct_coords[i])* diff(coordinate_system[k%],ct_coords[j]),k%,1,dim), if i#j then lg[j,i]:lg[i,j] ) ) ) else error("Unknown coordinate system type!"), if member('cylindrical,extra_args) then ( if member('z, ct_coords) then error("Cannot add Z coordinate to this metric, as it already exists") else ( dim:dim+1, ct_coords:append(ct_coords,['z]), if cframe_flag then ( fri:addrow(addcol(fri,zeromatrix(dim-1,1)),ident(dim)[dim]), lfg:addrow(addcol(lfg,zeromatrix(dim-1,1)),ident(dim)[dim]) ) else lg:addrow(addcol(lg,zeromatrix(dim-1,1)),ident(dim)[dim]) ) ) else if member('minkowski,extra_args) then ( if member('ct, ct_coords) then error("Cannot add CT coordinate to this metric, as it already exists") else ( dim:dim+1, /* ct_coords:append(ct_coords,['ct]),*/ ct_coords:cons('ct,ct_coords), if cframe_flag then ( /* fri:addrow(addcol(fri,zeromatrix(dim-1,1)),ident(dim)[dim]), lfg:addrow(addcol(lfg,zeromatrix(dim-1,1)),-ident(dim)[dim])*/ fri:apply(matrix,cons(-ident(dim)[1], makelist(transpose(apply(matrix,cons(zeromatrix(dim-1,dim-1)[1], makelist(transpose(fri)[i],i,1,dim-1))))[j],j,1,dim-1))), lfg:apply(matrix,cons(-ident(dim)[1], makelist(transpose(apply(matrix,cons(zeromatrix(dim-1,dim-1)[1], makelist(transpose(lfg)[i],i,1,dim-1))))[j],j,1,dim-1))) ) /* else lg:addrow(addcol(lg,zeromatrix(dim-1,1)),-ident(dim)[dim])*/ else lg:apply(matrix,cons(-ident(dim)[1], makelist(transpose(apply(matrix,cons(zeromatrix(dim-1,dim-1)[1], makelist(transpose(lg)[i],i,1,dim-1))))[j],j,1,dim-1))) ) ), if member('all,extra_args) then ( cmetric(), christof(false) ), if verbose then ( ldisplay(dim), ldisplay(ct_coords), if cframe_flag then ( ldisplay(lfg), ldisplay(fri) ) else ldisplay(lg) ), done )$ /*********************************** THE TENSORS OF GENERAL RELATIVITY ***********************************/ /* Compute the Ricci rotation coefficients in a frame base */ trrc(disp):=block ( [l,a,b,c,d], if not symmetricp(_lg(),dim) then error("Frame base calculations require a symmetric metric."), for a thru dim do for b thru dim do for c from b thru dim do if b=c then l[a,b,c]:0 else ( /* Landau-Lifshitz II p98, pp379 (Hungarian edition) */ l[a,b,c]:sum(sum((diff(fri[a][i%],ct_coords[k%])- diff(fri[a][k%],ct_coords[i%]))*fr[b][i%]*fr[c][k%], k%,1,dim),i%,1,dim), if ctrgsimp then l[a,b,c]:trigsimp(l[a,b,c]) else if ratchristof then l[a,b,c]:ratsimp(l[a,b,c]), l[a,c,b]:-l[a,b,c], if disp#false and debugmode then ldisplay(l[a,b,c]) ), for c thru dim do for a thru dim do for b from a thru dim do if a=b then lcs[c,a,b]:0 else ( lcs[c,b,a]:(l[a,b,c]+l[b,c,a]-l[c,a,b])/2, if ctrgsimp then lcs[c,b,a]:trigsimp(lcs[c,b,a]) else if ratchristof then lcs[c,b,a]:ratsimp(lcs[c,b,a]), if ratfac then lcs[c,b,a]:factor(lcs[c,b,a]), lcs[c,a,b]:-lcs[c,b,a] ), if disp=all or disp=lcs then for a thru dim do for b thru dim-1 do for c from b+1 thru dim do if lcs[a,b,c]#0 then ldisplay(lcs[a,b,c]) ); /* Compute the Christoffel-symbols, or, if a frame base is used, the Ricci rotation coefficients and place them in the arrays lcs[] and mcs[] */ christof(dis):=block ( [symmmetric:symmetricp(_lg(),dim)], if cframe_flag then ( if not symmmetric then error("Frame base calculations require a symmetric metric."), trrc(dis) ), if (not symmmetric and dis = lcs) then error("Christoffel symbols of the first kind require a symmetric metric"), if (not symmmetric and ctorsion_flag) then error("Torsion requires a symmetric metric"), if (not symmmetric and cnonmet_flag) then error("Nonmetricity requires a symmetric metric"), if ctorsion_flag and not member('kt,arrays) then contortion(tr), if cnonmet_flag and not member('nmc,arrays) then nonmetricity(nm), if symmmetric then ( for i thru dim do ( for j from i thru dim do ( if not cframe_flag then for k thru dim do ( lcs[i,j,k]:(diff(lg[j,k],ct_coords[i]) +diff(lg[k,i],ct_coords[j])-diff(lg[i,j],ct_coords[k]))/2, lcs[j,i,k]:lcs[i,j,k] ), for k thru dim do ( mcs[i,j,k]:ctaylor ( if diagmetric then ratsimp(_ug()[k,k]*lcs[i,j,k]) else sum(_ug()[l%,k]*lcs[i,j,l%],l%,1,dim) ), if ratchristof then mcs[i,j,k]:ratsimp(mcs[i,j,k]), mcs[j,i,k]:mcs[i,j,k], if ctorsion_flag then ( mcs[i,j,k]:mcs[i,j,k]-kt[i,j,k], if j#i then mcs[j,i,k]:mcs[j,i,k]-kt[j,i,k] ), if cnonmet_flag then ( mcs[i,j,k]:mcs[i,j,k]-nmc[i,j,k], if j#i then mcs[j,i,k]:mcs[j,i,k]-nmc[j,i,k] ) ) ) ) ) else block ( /* This is a direct solution of the field equations, g_{ab,c}-g_{bd}Gamma^d_{ac}-g_{dc}Gamma^d_{ba}=0. See, e.g., Moffat & Boal, Phys Rev D V11 N6 p1375, Eq 2.5 */ [_mcs:zeromatrix(dim^3,dim^3),_ucs,_abc,_gam,a,b,c,d], for a thru dim do for b thru dim do for c thru dim do for d thru dim do ( _mcs[(a-1)*dim^2+(b-1)*dim+c,(a-1)*dim^2+(c-1)*dim+d]:_mcs[(a-1)*dim^2+(b-1)*dim+c,(a-1)*dim^2+(c-1)*dim+d]+lg[b,d], _mcs[(a-1)*dim^2+(b-1)*dim+c,(b-1)*dim^2+(a-1)*dim+d]:_mcs[(a-1)*dim^2+(b-1)*dim+c,(b-1)*dim^2+(a-1)*dim+d]+lg[d,c] ), _ucs:_mcs^^(-1), if ratchristof then _ucs:ratsimp(_ucs), _abc:zeromatrix(dim^3,1), for a thru dim do for b thru dim do for c thru dim do ( _abc[(a-1)*dim^2+(b-1)*dim+c,1]:diff(lg[b,c],ct_coords[a]), if ratchristof then _abc[(a-1)*dim^2+(b-1)*dim+c,1]:ratsimp(_abc[(a-1)*dim^2+(b-1)*dim+c,1]) ), _gam:_ucs._abc, if ratchristof then _gam:ratsimp(_gam), for a thru dim do for b thru dim do for c thru dim do mcs[a,b,c]:_gam[(a-1)*dim^2+(b-1)*dim+c][1] ), if dis = all or dis = lcs then for i thru dim do for j from i thru dim do for k thru dim do if lcs[i,j,k]#0 then ldisplay(lcs[i,j,k]), if dis = mcs or dis = all then for i thru dim do for j from (if symmmetric then i else 1) thru dim do for k thru dim do if mcs[i,j,k]#0 then ldisplay(mcs[i,j,k]), done )$ /* In a moving frame, the Ricci-tensor is computed from the covariant curvature tensor */ tricci(disp):=block ( [df], if not symmetricp(_lg(),dim) then error("Frame base calculations require a symmetric metric."), if not member('lriem,arrays) then triemann(false), kill(ric), df:true, for i thru dim do for j from i thru dim do ( ric[i,j]:sum(sum(ufg[k%,l%]*lriem[l%,k%,i,j],k%,1,dim),l%,1,dim), if ctrgsimp then ric[i,j]:trigsimp(ric[i,j]) else if ratfac then ric[i,j]:factor(ric[i,j]) else ric[i,j]:ctaylor(ratsimp(ric[i,j])), if i#j then ric[j,i]:ric[i,j], if disp and ric[i,j]#0 then ( ldisplay(ric[i,j]), df:false ) ), tracer:sum(sum(ric[i%,j%]*ufg[i%,j%],i%,1,dim),j%,1,dim), if df and disp then print("THIS SPACETIME IS EMPTY AND/OR FLAT"), done ); /* In case of a holonomic metric, the Christoffel-symbols of the second kind are used. */ ricci(dis):=if cframe_flag then tricci(dis) else block ( [suma,sumb,flat:true,symmmetric:symmetricp(_lg(),dim)], modedeclare(flat,boolean), if not member('mcs,arrays) then christof(false), if symmmetric then for i thru dim do ( for j from i thru dim do ( suma:0,sumb:0, for k thru dim do ( if k#i then suma:suma+(diff(mcs[j,i,k],ct_coords[k]) -diff(mcs[j,k,k],ct_coords[i])), sumb:sumb+sum(mcs[l%,k,k]*mcs[i,j,l%]-mcs[k,i,l%]*mcs[j,l%,k],l%,1,dim) ), ric[i,j]:ctaylor(suma+sumb), if ratfac then ric[i,j]:factor(ctaylor(ratsimp(ric[i,j]))), ric[j,i]:ric[i,j] ) ) else ( for m thru dim do for n thru dim do ric[m,n]:ctaylor(sum(diff(mcs[m,n,a],ct_coords[a])- (diff(mcs[m,a,a]+mcs[a,m,a],ct_coords[n])+ diff(mcs[n,a,a]+mcs[a,n,a],ct_coords[m]))/4+ sum(-mcs[m,o,a]*mcs[a,n,o]+mcs[m,n,a]*mcs[a,o,o],o,1,dim) ,a,1,dim)), if ratfac then ric[i,j]:factor(ctaylor(ratsimp(ric[i,j]))) ), if dis#false then ( for i thru dim do for j from (if symmmetric then i else 1) thru dim do if ric[i,j]#0 then ( flat:false, ldisplay(ric[i,j]) ), if flat then print("THIS SPACETIME IS EMPTY AND/OR FLAT") ), tracer:if diagmetric then ctaylor(ratsimp(sum(ric[i%,i%]*ug[i%,i%],i%,1,dim))) else ctaylor(sum(sum(ric[i%,j%]*ug[i%,j%],i%,1,dim),j%,1,dim)), done )$ /* Computing the mixed Ricci-tensor */ uricci(dis):=block( [flat:true], modedeclare(flat,boolean), if not (member('ric,arrays) or matrixp(ric)) then ricci(false), for i thru dim do for j thru dim do uric[i,j]:0, for i thru dim do ( for j thru dim do ( uric[i,j]:if diagmetric then ratsimp(ctaylor(ric[i,j]*_ug()[j,j])) else ctaylor(sum(ric[i,k%]*_ug()[k%,j],k%,1,dim)), if ratfac then uric[i,j]:factor(ctaylor(ratsimp(uric[i,j]))) ) ), if dis#false then ( for i thru dim do for j thru dim do if uric[i,j]#0 then ( flat:false, ldisplay(uric[i,j]) ), if flat then print("THIS SPACETIME IS EMPTY AND/OR FLAT") ), done )$ /* The scalar curvature */ scurvature():= ( if not (member('uric,arrays) or matrixp(uric)) then uricci(false), if ratfac then factor(tracer) else tracer )$ /* Computing the mixed Einstein-tensor */ einstein(dis):=block ( [flat:true], modedeclare(flat,boolean), if not (member('uric,arrays) or matrixp(uric)) then uricci(false), for i thru dim do for j thru dim do ( if i=j then ein[i,j]:ctaylor(uric[i,j]-tracer/2) else ein[i,j]:ctaylor(uric[i,j]), if ratfac then ein[i,j]:factor(ratsimp(ein[i,j])) else if rateinstein then ein[i,j]:ratsimp(ein[i,j]), if dis#false and ein[i,j]#0 then ( flat:false, ldisplay(ein[i,j]) ) ), if dis#false and flat then print("THIS SPACETIME IS EMPTY AND/OR FLAT"), done )$ /* Computing the covariant Einstein-tensor */ leinstein(dis):=block ( if not (member('ein,arrays) or matrixp(ein)) then einstein(false), for i thru dim do for j thru i do ( lein[i,j]:if diagmetric then ein[i,j]*lg[j,j] else sum(ein[i,k%]*lg[k%,j],k%,1,dim), if i#j then lein[j,i]:lein[i,j], if dis#false and lein[i,j]#0 then ldisplay(lein[i,j]) ) )$ /* The Riemann-tensor is computed so that its indices are compatible with the index ordering in the itensor package: l _l _l _l _m _l _m R = | - | + | | - | | ijk ij,k ik,j mk ij mj ik This index ordering is somewhat unusual. In this form, the (3,1) Riemann tensor is antisymmetric in its second and third indices. The computation of the Weyl tensor follows the same index ordering. */ /* This function computes the (1,3) Riemann tensor */ riemann(dis):=if cframe_flag then block ( if not member('lriem,arrays) then triemann(false), for i thru dim do for j thru dim do for k thru dim do for l thru dim do riem[i,j,k,l]:sum(lriem[i,j,k,m%]*ufg[l,m%],m%,1,dim), done ) else block ( [flat : true], modedeclare(flat,boolean), if not member('mcs,arrays) then christof(false), for i thru dim do for j thru dim do for k thru dim do for l thru dim do riem[i,j,k,l]:0, for i thru dim do for j thru dim do for k thru j do for l thru dim do ( riem[i,j,k,l]:ctaylor(diff(mcs[i,j,l], ct_coords[k]) - diff(mcs[i,k,l], ct_coords[j]) + sum(mcs[m%,k,l]*mcs[i,j,m%] - mcs[m%,j,l]*mcs[i,k,m%], m%, 1, dim)), if ratfac then riem[j,i,k,l] : factor(ctaylor(ratsimp(riem[j,i,k,l]))) else if ratriemann then riem[j,i,k,l] : ctaylor(ratsimp(riem[j,i,k,l])), if k # j then riem[i,k,j,l] : -riem[i,j,k,l], if dis#false and riem[i,j,k,l] # 0 then (flat : false, ldisplay(riem[i,j,k,l])) ), if dis#false and flat then print("This spacetime is flat"), done )$ /* Computing the covariant Riemann-tensor in a moving frame */ triemann(disp):=block ( [df], if not symmetricp(_lg(),dim) then error("Frame base calculations require a symmetric metric."), if not member('lcs,arrays) then trrc(false), kill(lriem), df:true, for i thru dim do for j from i thru dim do for k thru dim do for l from k thru dim do ( if i=j or k=l then lriem[l,j,i,k]:0 /* Wald (1984), p51, eq. 3.4.21 */ else lriem[l,j,i,k]:ctaylor(sum(fr[i][a%]*diff(lcs[j,k,l],ct_coords[a%])- fr[j][a%]*diff(lcs[i,k,l],ct_coords[a%]),a%,1,dim)-sum(sum(ufg[a%,b%]* (lcs[i,b%,k]*lcs[j,a%,l]-lcs[j,b%,k]*lcs[i,a%,l]+lcs[i,b%,j]*lcs[a%,k,l]- lcs[j,b%,i]*lcs[a%,k,l]),b%,1,dim),a%,1,dim)), if ctrgsimp then lriem[l,j,i,k]:trigsimp(ev(lriem[l,j,i,k])) else if ratriemann then lriem[l,j,i,k]:ratsimp(ev(lriem[l,j,i,k])) else lriem[l,j,i,k]:ev(lriem[l,j,i,k]), if ratfac then lriem[l,j,i,k]:factor(lriem[l,j,i,k]), lriem[l,i,j,k]:-lriem[l,j,i,k], lriem[k,j,i,l]:-lriem[l,j,i,k], lriem[k,i,j,l]:lriem[l,j,i,k], if disp and lriem[l,j,i,k]#0 then ( ldisplay(lriem[l,j,i,k]), df:false ) ), if df and disp then print("THIS SPACETIME IS FLAT"), done ); /* This function computes the covariant Riemann tensor */ lriemann(dis):=if cframe_flag then triemann(dis) else block ( [symmmetric:symmetricp(_lg(),dim)], if not member('riem,arrays) then riemann(false), for i thru dim do for j thru dim do for k thru dim do for l thru dim do lriem[i,j,k,l]:0, for i thru dim do for j thru dim do for k thru (if symmmetric then j else dim) do for l thru (if symmmetric then i else dim) do ( lriem[i,j,k,l] : if diagmetric then ratsimp(riem[i,j,k,l]*lg[l,l]) else sum(riem[i,j,k,m%]*lg[m%,l], m%, 1, dim), if ratriemann then lriem[i,j,k,l] : ratsimp(lriem[i,j,k,l]), if symmmetric then ( if j # k then lriem[i,k,j,l] : -lriem[i,j,k,l], if l # i then ( lriem[l,j,k,i] : -lriem[i,j,k,l], if j # k then lriem[l,k,j,i] : lriem[i,j,k,l] ) ), if dis#false and lriem[i,j,k,l] # 0 then ldisplay(lriem[i,j,k,l]) ), done )$ /* This function computes the contravariant Riemann tensor */ uriemann(dis):=block ( [symmmetric:symmetricp(_lg(),dim)], if not member('riem,arrays) then riemann(false), for i thru dim do for j thru dim do for k thru dim do for l thru dim do uriem[i,j,k,l]:0, for i thru dim do for j thru dim do for k thru (if symmmetric then j else dim) do for l thru (if symmmetric then i else dim) do ( uriem[i,j,k,l] : if diagmetric then ratsimp(riem[i,j,k,l]*_ug()[i,i]*_ug()[j,j]*_ug()[k,k]) else sum(sum(sum( riem[m%,n%,o%,l]*_ug()[i,m%]*_ug()[j,n%]*_ug()[k,o%], m%, 1, dim), n%, 1, dim), o%, 1, dim), if ratriemann then uriem[i,j,k,l] : ratsimp(uriem[i,j,k,l]), if symmmetric then ( if j # k then uriem[i,k,j,l] : -uriem[i,j,k,l], if l # i then ( uriem[l,j,k,i] : -uriem[i,j,k,l], if j # k then uriem[l,k,j,i] : uriem[i,j,k,l] ) ), if dis#false and uriem[i,j,k,l] # 0 then ldisplay(uriem[i,j,k,l]) ), done )$ /* The Kretchmann invariant is R^2=lriem[i,j,k,l]*uriem[i,j,k,l] */ rinvariant():=kinvariant:sum(sum(sum(sum(lriem[i%,j%,k%,l%]*uriem[i%,j%,k%,l%], i%,1,dim),j%,1,dim),k%,1,dim),l%,1,dim)$ /* The covariant Weyl tensor. See the note on the Riemann tensor above for index ordering. */ weyl(dis):=block ( [flat:true,symmmetric:symmetricp(_lg(),dim)], modedeclare(flat,boolean), if dim=2 then ( print("ALL 2 DIMENSIONAL SPACETIMES ARE CONFORMALLY FLAT"), return(done) ), if not ((member('ric,arrays) or matrixp(ric)) and member('lriem,arrays)) then ( ricci(false), riemann(false), lriemann(false) ), for i thru dim do for j thru dim do for k thru dim do for l thru dim do weyl[i,j,k,l]:0, for i thru dim do for j from (if symmmetric then i+1 else 1) thru dim do for k from (if symmmetric then i else 1) thru dim do for l from (if symmmetric then k+1 else 1) thru (if (symmmetric and k=i) then j else dim) do ( weyl[l,j,i,k]:ctaylor(lriem[l,j,i,k]+ 1/(dim-2)*(_lg()[i,l]*ric[j,k]-_lg()[i,k]*ric[l,j] +_lg()[j,k]*ric[l,i]-_lg()[j,l]*ric[k,i])- tracer/((dim-1)*(dim-2))* (_lg()[i,l]*_lg()[k,j]-_lg()[i,k]*_lg()[l,j]) ), if ratfac then weyl[l,j,i,k]:factor(ratsimp(weyl[l,j,i,k])) else if ratweyl then weyl[l,j,i,k]:ratsimp(weyl[l,j,i,k]), if symmmetric then ( weyl[k,i,j,l]:weyl[l,j,i,k], weyl[k,j,i,l]:weyl[l,i,j,k]:-weyl[l,j,i,k], if i#k or j>l then ( weyl[j,l,k,i]:weyl[i,k,l,j]:weyl[l,j,i,k], weyl[j,k,l,i]:weyl[i,l,k,j]:-weyl[l,j,i,k] ) ), if dis#false and weyl[l,j,i,k]#0 then ( flat:false, ldisplay(weyl[l,j,i,k]) ) ), if dis#false and flat then print("THIS SPACETIME IS CONFORMALLY FLAT"), done )$ /*********************************** MISCELLANEOUS CALCULATIONS ***********************************/ /* The geodesic equation of motion */ cgeodesic(dis):=block ( [s,lgeod], map(lambda([u],apply('remove,[u,constant])),ct_coords), depends(ct_coords,s), for i thru dim do lgeod[i]:if diagmetric then ratsimp(lg[i,i]*diff(ct_coords[i],s,2)+ sum(diff(lg[i,i],ct_coords[a%])*diff(ct_coords[i],s)* diff(ct_coords[a%],s),a%,1,dim)- 1/2*sum(diff(lg[a%,a%],ct_coords[i])* diff(ct_coords[a%],s)^2,a%,1,dim)) else ratsimp(sum(lg[i,a%]*diff(ct_coords[a%],s,2),a%,1,dim)+ sum(sum(diff(lg[i,b%],ct_coords[a%])*diff(ct_coords[b%],s)* diff(ct_coords[a%],s)- 1/2*diff(lg[a%,b%],ct_coords[i])*diff(ct_coords[a%],s)* diff(ct_coords[b%],s),a%,1,dim),b%,1,dim)), for i thru dim do ( geod[i]:factor(ratexpand(sum(ratsimp(_ug()[i,a%]*lgeod[a%]),a%,1,dim))), if dis#false then ldisplay(geod[i]) ), done )$ /* The 4-dimensional Brans-Dicke vacuum field equations are represented by the array bd and the scalar equation is generated by setting the d'Alembertian of the scalar field to zero. That is, one calls the function dscalar on the scalar field. */ bdvac(zz):=block ( [addd,boxq:0], if dim#4 then error("This program is restricted to 4 dimensions"), for i:1 thru 4 do for j:1 thru 4 do addd[i,j]:w/zz^2*(diff(zz,ct_coords[i])*diff(zz,ct_coords[j])-lg[i,j]* sum(diff(zz,ct_coords[k%])*diff(zz,ct_coords[k%])*_ug()[k%,k%],k%,1,4)/2)+ (diff(diff(zz,ct_coords[i]),ct_coords[j])-sum(mcs[i,j,k%]* diff(zz,ct_coords[k%]),k%,1,4)-lg[i,j]*boxq)/zz, for i:1 thru 4 do for j:1 thru 4 do bd[i,j]:ratsimp(ric[i,j]-r*lg[i,j]/2-0*t[i,j]-addd[i,j]), remarray(addd), done )$ /* Compute the Euler-Lagrange equations for the density of the invariant R^2. The form is (D is the Kronecker-delta): b 2 b b bc (1/2)*D R - 2 R R + 2 D []R -2 g R a a a ;ac The equations are sometimes less complex if tracer is given a parametric name with dependencies corresponding to those of the scalar curvature. */ invariant1():=block ( for aa thru dim do for bb thru dim do ( inv1[aa,bb]:0 ), if diagmetric then for aa thru dim do for bb thru dim do ( inv1[aa,bb]:ratsimp(kdelt(aa,bb)*tracer^2/2-2*tracer*uric[aa,bb]- 2*kdelt(aa,bb)*dscalar(tracer)+2*_ug()[aa,aa]* (diff(diff(tracer,ct_coords[bb]),ct_coords[aa])- sum(mcs[bb,aa,k%]*diff(tracer,ct_coords[k%]),k%,1,dim))) ) else for aa thru dim do for bb thru dim do ( inv1[aa,bb]:ratsimp(kdelt(aa,bb)*tracer^2/2-2*tracer*uric[aa,bb]- 2*kdelt(aa,bb)*dscalar(tracer)+2*sum(_ug()[bb,c%]* (diff(diff(tracer,ct_coords[aa]),ct_coords[c%])-sum( mcs[aa,c%,k%]*diff(tracer,ct_coords[k%]),k%,1,dim)),c%,1,dim)) ) )$ invariant2():="NOT YET IMPLEMENTED"; bimetric():="NOT YET IMPLEMENTED"; /* Compute a Newman-Penrose null tetrad */ nptetrad(disp):=block( [nu], if cframe_flag#true then error("Frame base required"), if dim#4 then error("This calculation requires dim=4"), nu:ident(4), nu[1,1]:-1, if lfg#nu then error("Frame metric signature of (-,+,+,+) required"), np:ident(4), np[1]:(fri[1]+fri[2])/sqrt(2), np[2]:(fri[1]-fri[2])/sqrt(2), np[3]:(fri[3]+%i*fri[4])/sqrt(2), np[4]:(fri[3]-%i*fri[4])/sqrt(2), if ctrgsimp then np:trigsimp(np) else np:ratsimp(np), if ratfac then np:factor(np), if disp then ldisplay(np), npi:np.ug, if ctrgsimp then npi:trigsimp(npi) else npi:ratsimp(npi), if ratfac then npi:factor(npi), if disp then ldisplay(npi), done ); contract4(w,l1,l2,l3,l4)::= ( if dim#4 then error("This calculation requires dim=4"), modedeclare([l1,l2,l3,l4],fixnum), block ( [ans:0,i,j,k,l], for i thru 4 do for j thru 4 do for k thru 4 do for l thru 4 do ans:ans+w[l,j,i,k]*npi[l1,i]*npi[l2,j]*npi[l3,k]*npi[l4,l], if ctrgsimp then ans:trigsimp(ans) else ratsimp(ans) ) ); /* Compute the Newman-Penrose coefficients */ psi(disp):=block ( /* To compute these values, we must transform the Weyl-tensor to coordinate representation if we've been working in a frame field. */ [w], if cframe_flag then ( for i thru dim do for j thru dim do for k thru dim do for l thru dim do w[i,j,k,l]:0, for i thru dim do for j from i+1 thru dim do for k from i thru dim do for l from k+1 thru (if k=i then j else dim) do ( w[l,j,i,k]:sum(sum(sum(sum(weyl[l%,j%,i%,k%]*fri[i%,i]*fri[j%,j]* fri[k%,k]*fri[l%,l],i%,1,dim),j%,1,dim),k%,1,dim),l%,1,dim), if ratfac then w[l,j,i,k]:factor(ratsimp(w[l,j,i,k])) else if ratweyl then w[l,j,i,k]:ratsimp(w[l,j,i,k]), w[k,i,j,l]:w[l,j,i,k], w[k,j,i,l]:w[l,i,j,k]:-w[l,j,i,k], if i#k or j>l then ( w[j,l,k,i]:w[i,k,l,j]:w[l,j,i,k], w[j,k,l,i]:w[i,l,k,j]:-w[l,j,i,k] ) ) ) else w:'weyl, psi[0]:ratsimp(-contract4(w,1,3,1,3)), if disp then ldisplay(psi[0]), psi[1]:ratsimp(-contract4(w,1,2,1,3)), if disp then ldisplay(psi[1]), psi[2]:ratsimp(gfactor(-1/2*(contract4(w,1,2,1,2)+contract4(w,1,2,3,4)))), if disp then ldisplay(psi[2]), psi[3]:ratsimp(contract4(w,1,2,2,4)), if disp then ldisplay(psi[3]), psi[4]:ratsimp(-contract4(w,2,4,2,4)), if disp then ldisplay(psi[4]), done ); /* Compute the Petrov-classification of the metric; algorithm based on Classifying geometries in general relativity: III Classification in practice by Pollney, Skea, d'Inverno, Class. Quant. Grav. 17 2885-2992 (2000) The algoritm has been optimized to compute only quantities that are required for a specific (sub)case. As of 12/19/2004, it is essentially untested, may even contain typos in some expressions. */ petrov():=block ( [ T:[ 0,'N,'II,'III, 'D,'II,'II, 7, 'II,'I, 'I, 11,'II, 13, 14,15, 'N,'I, 'I, 19,'II, 21, 13,23, 'III,19, 11, 27, 7, 23, 15,31], P:1 ], if psi[4]#0 then P:P+1, if psi[3]#0 then P:P+2, if psi[2]#0 then P:P+4, if psi[1]#0 then P:P+8, if psi[0]#0 then P:P+16, if numberp(T[P]) and T[P]>0 then ( if T[P]=7 then if ratsimp(psi[3]^2-3*psi[2]*psi[4])=0 then 'D else 'II else if T[P]=11 then if ratsimp(27*psi[4]^2*psi[1]+64*psi[3]^3)=0 then 'II else 'I else if T[P]=13 then if ratsimp(psi[1]^2*psi[4]+2*psi[2]^3)=0 then 'II else 'I else if T[P]=14 then if ratsimp(9*psi[2]^2-16*psi[1]*psi[3])=0 then 'II else 'I else if T[P]=15 then if ratsimp(3*psi[2]^2-4*psi[1]*psi[3])=0 and ratsimp(psi[2]*psi[3]-3*psi[1]*psi[4])=0 then 'II else 'I else if T[P]=19 then if ratsimp(psi[0]*psi[4]^3-27*psi[3]^4)=0 then 'II else 'I else if T[P]=21 then if ratsimp(9*psi[2]^2-psi[4]^2)=0 then 'D else 'I else if T[P]=23 then block ( [I:ratsimp(psi[0]*psi[4]+3*psi[2]^2)], if I=0 and ratsimp(4*psi[2]*psi[4]-3*psi[3]^2)=0 then 'III else block ( [J:ratsimp(4*psi[2]*psi[4]-3*psi[3]^2)], if ratsimp(psi[4]*I^2-3*J*(psi[0]*J-2*psi[2]*I))=0 then 'II else 'I ) ) else if T[P]=27 then if ratsimp(psi[0]*psi[3]^2-psi[1]^2*psi[4])=0 then if ratsimp(psi[0]*psi[4]+2*psi[1]*psi[3])=0 then 'D else if ratsimp(psi[0]*psi[4]-16*psi[1]*psi[3])=0 then 'II else 'I else block ( [I:ratsimp(psi[0]*psi[4]+2*psi[1]*psi[3])], if I=0 then block ( [J:ratsimp(-psi[0]*psi[3]^2-psi[1]^2*psi[4])], if J=0 then 'III else if ratsimp(I^3-27*J^2)=0 then 'II else 'I ) else 'I ) else block ( [H:ratsimp(psi[0]*psi[2]-psi[1]^2)], if H=0 then if ratsimp(psi[0]*psi[3]-psi[1]*psi[2])=0 then if ratsimp(psi[0]*psi[4]-psi[2]^2)=0 then 'N else 'I else block ( [E:ratsimp(psi[0]*psi[4]-psi[2]^2)], if E=0 then if ratsimp(37*psi[2]^2+27*psi[1]*psi[3])=0 then 'II else 'I else block ( [ A:ratsimp(psi[1]*psi[3]+psi[2]^2), I:ratsimp(E-4*A) ], if I#0 and ratsimp(I^3-27*(psi[4]*H-psi[3]^2*psi[0]+ psi[1]*psi[2]*psi[3]+psi[2]*A)^2)=0 then 'II else 'I ) ) else block ( [I:ratsimp(psi[0]*psi[4]-psi[2]^2-4*(psi[1]*psi[3]+psi[2]^2))], if I=0 then if ratsimp(psi[4]*H-psi[3]^2*psi[0]+psi[1]*psi[2]*psi[3]+ psi[2]*(psi[1]*psi[3]+psi[2]^2))=0 then 'III else 'I else if ratsimp(psi[0]^2*psi[3]-psi[0]*psi[1]*psi[2]-2*psi[1]*H)=0 then if ratsimp(psi[0]^2*I-12*H^2)=0 then 'D else if ratsimp(psi[0]^2*I-3*H^2)=0 then 'II else 'I else block ( [J:ratsimp(psi[4]*H-psi[3]^2*psi[0]+psi[1]*psi[2]*psi[3]+ psi[2]*(psi[1]*psi[3]+psi[2]^2))], if J#0 and ratsimp(I^3-27*J^2)=0 then 'II else 'I ) ) ) ) else T[P] ); /* Old algorithm by anonymous, intended for diagonal metrics petrov():=block ( [ii,jj,gg,hh], ii:ratsimp(psi[0]*psi[4]-4*psi[1]*psi[3]+3*psi[2]^2), jj:ratsimp(determinant(matrix([psi[0],psi[1],psi[2]], [psi[1],psi[2],psi[3]], [psi[2],psi[3],psi[4]]))), if ratsimp(ii^3-27*jj^2)#0 then "Type is I" else ( gg:ratsimp(psi[0]^2*psi[3]-3*psi[0]*psi[1]*psi[2]+2*psi[1]^3), if ii=0 and jj=0 then ( if gg=0 then ( hh:ratsimp(psi[0]*psi[2]-psi[1]^2), if hh=0 then ( if psi[0] = 0 and psi[1] = 0 and psi[2] = 0 and psi[3] = 0 and psi[4] = 0 then "Type is FLAT(0)" else "Type is N" ) else "Type is III" ) else "Type is III" ) else ( hh:ratsimp(psi[0]*psi[2]-psi[1]^2), if ratsimp(gg-(psi[0]^2*ii-12*hh^2))#0 then "Type is II" else "Type is D" ) ) )$*/ frame_bracket(fr,fri,diagframe):=block ( for a thru dim do for b thru dim do for c thru dim do fb[a,b,c]:sum(sum(fr[a,e%]*(diff(fri[e%,c],ct_coords[f%])- diff(fri[f%,c],ct_coords[e%]))*fr[b,c],e%,1,dim),f%,1,dim) ); /* The findde function finds a list of unique differential equations in the list or 2-dimensional array a. The variable deindex is a global list of the indices corresponding to these unique differential equations. */ findde(a,n):= ( modedeclare(n,fixnum), if n=1 then findde1(a) else if n=2 then findde2(a) else if n=3 then findde3(a) else error("Invalid dimension:",n) )$ findde1(list):=block ( [inflag:true,deriv:nounify('derivative),l:[],l1,q], deindex:[], for i while list#[] do ( l1:factor(num(first(list))), list:rest(list), if not freeof(deriv,l1) then ( deindex:cons(i,deindex), if inpart(l1,0)#"*" then l:cons(l1,l) else ( q:1, for j thru length(l1) do if not freeof(deriv,inpart(l1,j)) then q:q*inpart(l1,j), l:cons(q,l) ) ) ), cleanup(l) )$ findde2(a):=block ( [inflag:true,deriv:nounify('derivative),l:[],t,q], deindex:[], for i thru dim do for j thru dim do ( t:factor(num(a[i,j])), if not freeof(deriv,t) then ( deindex:cons([i,j],deindex), if inpart(t,0)#"*" then l:cons(t,l) else ( q:1, for n thru length(t) do if not freeof(deriv,inpart(t,n)) then q:q*inpart(t,n), l:cons(q,l) ) ) ), cleanup(l) )$ findde3(a):=block ( [inflag:true,deriv:nounify('derivative),l:[],t,q], deindex:[], for i thru dim do for j thru dim do for k thru dim do ( t:factor(num(a[i,j,k])), if not freeof(deriv,t) then ( deindex:cons([i,j,k],deindex), if inpart(t,0)#"*" then l:cons(t,l) else ( q:1, for n thru length(t) do if not freeof(deriv,inpart(t,n)) then q:q*inpart(t,n), l:cons(q,l) ) ) ), cleanup(l) )$ cleanup(ll):=block ( [a,l:[],index:[]], while ll#[] do ( a:first(ll), ll:rest(ll), if not member(a,ll) then ( l:cons(a,l), index:cons(first(deindex),index) ), deindex:rest(deindex) ), deindex:index, l )$ /*********************************** AUXILIARY FUNCTIONS (not used in ctensor) ***********************************/ /* Covariant and contravariant gradients */ cograd(f,xyz):=block ( [], for mm thru dim do ( arraysetapply(xyz,[mm],ratsimp(diff(f,ct_coords[mm]))) ) )$ contragrad(f,xyz):=block ( if diagmetric then for mm thru dim do arraysetapply(xyz,[mm],ratsimp(ratsimp(_ug()[mm,mm]*diff(f,ct_coords[mm])))) else for mm thru dim do arraysetapply(xyz,[mm],sum(_ug()[m,n]*diff(f,ct_coords[n%]),n%,1,dim)) )$ /* The d'Alembertian of a scalar */ dscalar(phi):=block ( if diagmetric then ratsimp(1/sqrt(-gdet)*sum(diff(_ug()[i%,i%]* sqrt(-gdet)*diff(phi,ct_coords[i%]),ct_coords[i%]),i%,1,dim)) else ratsimp(1/sqrt(-gdet)*sum(sum(diff(_ug()[i%,j%]*sqrt(-gdet)* diff(phi,ct_coords[j%]),ct_coords[i%]),i%,1,dim),j%,1,dim)) )$ /* Covariant divergence. The object gxyz must be a mixed 2nd rank tensor whose first index is covariant. */ checkdiv(gxyz):=block ( modedeclare([i,j%],fixnum), if diagmatrixp(gxyz,dim) then for i thru dim do print(div[i]:ratsimp(ctaylor(1/sqrt(-gdet)*diff(sqrt(-gdet)*gxyz[i,i], ct_coords[i])-sum(mcs[i,j%,j%]*gxyz[j%,j%],j%,1,dim)))) else for i thru dim do print(div[i]:ratsimp(ctaylor(1/sqrt(-gdet)*sum(diff(sqrt(-gdet)*gxyz[i,j%], ct_coords[j%]),j%,1,dim)-sum(sum(mcs[i,j%,a%]*gxyz[a%,j%],a%,1,dim),j%,1,dim)))) )$ /* Contortion. Mostly generated with ic_convert(). */ contortion(tr):=block ( for i thru dim do for j thru dim do for k thru dim do kt[i,j,k]:sum(_ug()[k,l%]* (-sum(tr[l%,i,m%]*lg[j,m%],m%,1,dim)-sum(lg[l%,m%]*tr[i,j,m%],m%,1,dim)- sum(tr[l%,j,m%]*lg[i,m%],m%,1,dim)), l%,1,dim)/2 )$ /* Nonmetricity. Also generated mainly with ic_convert(). */ nonmetricity(nm):=block ( for i thru dim do for j thru dim do for k thru dim do nmc[i,j,k]:-nm[i]*kdelt(j,k)-kdelt(i,k)*nm[j]+ sum(_ug()[k,l%]*nm[l%],l%,1,dim)*lg[i,j] )$ /******************************************************************** * Untested code from miscellaneous files... /* The second covariant derivative of the covariant Ricci Tensor */ riccicov(ii,jj,kk,ll):=block( modedeclare([ii,jj,kk,ll,a%,b%],fixnum), if diagmetric then h[ii,jj,kk,ll]:h[jj,ii,kk,ll]: ratsimp(sum( mcs[a%,ii,jj]*ric[jj,jj]*mcs[kk,ll,a%]+ mcs[a%,jj,ii]*ric[ii,ii]*mcs[kk,ll,a%]+ ric[a%,a%]*mcs[ii,kk,a%]*mcs[jj,ll,a%]+ mcs[a%,kk,ii]*ric[ii,ii]*mcs[jj,ll,a%]+ ric[a%,a%]*mcs[ii,ll,a%]*mcs[jj,kk,a%]+ mcs[a%,kk,jj]*ric[jj,jj]*mcs[ii,ll,a%]- diff(ric[ii,jj],ct_coords[a%],1)*mcs[kk,ll,a%],a%,1,dim) -(diff(ric[ii,ii],ct_coords[kk],1)*mcs[jj,ll,ii]+ ric[ii,ii]*diff(mcs[jj,kk,ii],ct_coords[ll],1)+ diff(ric[ii,ii],ct_coords[ll],1)*mcs[jj,kk,ii]+ diff(ric[jj,jj],ct_coords[kk],1)*mcs[ii,ll,jj]+ ric[jj,jj]*diff(mcs[ii,kk,jj],ct_coords[ll],1)+ diff(ric[jj,jj],ct_coords[ll],1)*mcs[ii,kk,jj]) +diff(diff(ric[ii,jj],ct_coords[kk],1),ct_coords[ll],1)) else h[ii,jj,kk,ll]:h[jj,ii,kk,ll]: ratsimp(sum(sum( mcs[a%,ii,b%]*ric[b%,jj]*mcs[kk,ll,a%]+ mcs[a%,jj,b%]*ric[b%,ii]*mcs[kk,ll,a%]+ ric[a%,b%]*mcs[ii,kk,a%]*mcs[jj,ll,b%]+ mcs[a%,kk,b%]*ric[b%,ii]*mcs[jj,ll,a%]+ ric[a%,b%]*mcs[ii,ll,a%]*mcs[jj,kk,b%]+ mcs[a%,kk,b%]*ric[b%,jj]*mcs[ii,ll,a%],a%,1,dim),b%,1,dim) - sum( diff(ric[ii,jj],ct_coords[a%],1)*mcs[kk,ll,a%]+ diff(ric[a%,ii],ct_coords[kk],1)*mcs[jj,ll,a%]+ ric[a%,ii]*diff(mcs[jj,kk,a%],ct_coords[ll],1)+ diff(ric[a%,ii],ct_coords[ll],1)*mcs[jj,kk,a%]+ diff(ric[a%,jj],ct_coords[kk],1)*mcs[ii,ll,a%]+ ric[a%,jj]*diff(mcs[ii,kk,a%],ct_coords[ll],1)+ diff(ric[a%,jj],ct_coords[ll],1)*mcs[ii,kk,a%],a%,1,dim) +diff(diff(ric[ii,jj],ct_coords[kk],1),ct_coords[ll],1)))$ *******************************************************************/ put('ctensor,'v20041204,'version)$