module numz integer, parameter:: b4 = selected_real_kind(4) end module module maxsize integer nmax parameter (nmax=8000) end module module ccm_time_mod integer, parameter:: b8 = selected_real_kind(10) integer,save,private :: ccm_start_time(8) = (/-1, -1, -1, -1, -1, -1, -1, -1/) contains function ccm_time() implicit none real(b8) :: ccm_time,tmp integer,parameter :: norm(13)=(/ & 0, 2678400, 5097600, 7776000,10368000,13046400,& 15638400,18316800,20995200,23587200,26265600,28857600,31536000/) integer,parameter :: leap(13)=(/ & 0, 2678400, 5184000, 7862400,10454400,13132800,& 15724800,18403200,21081600,23673600,26352000,28944000,31622400/) integer :: values(8),m,sec call date_and_time(values=values) if(ccm_start_time(1) .eq. -1)ccm_start_time=values if(mod(values(1),4) .eq. 0)then m=leap(values(2)) else m=norm(values(2)) endif sec=((values(3)*24+values(5))*60+values(6))*60+values(7) tmp=real(m,b8)+real(sec,b8)+real(values(8),b8)/1000.0_b8 if(values(1) .ne. ccm_start_time(1))then if(mod(ccm_start_time(1),4) .eq. 0)then tmp=tmp+real(leap(13),b8) else tmp=tmp+real(norm(13),b8) endif endif ccm_time=tmp end function end module ccm_time_mod module special contains function j1(x) use numz implicit none real(b4) j1,x integer n real(b4) bjn,by,bin,bk n=1 call besse(x,bjn,by,bin,bk,n) j1=bjn end function function h0(x) use numz implicit none complex(b4) h0,j real(b4) x integer n real(b4) bjn,by,bin,bk j=cmplx(0,1) n=0 call besse(x,bjn,by,bin,bk,n) h0=bjn-j*by end function function h1(x) use numz implicit none complex(b4) h1,j real(b4) x integer n real(b4) bjn,by,bin,bk j=cmplx(0,1) n=1 call besse(x,bjn,by,bin,bk,n) h1=bjn-j*by end function subroutine besse (x, bjn, by, bin, bk,nord) use numz implicit none real(b4) x, bjn, by, bin, bk integer nord logical finished, done_wron,done real(b4) an, ai, ak, fn, gam, t, s, xa, c, d, phi real(b4) pn, qn, u, den, xb, pi, bi, xi real(b4) bj0, by0, bi0, bk0, bk1, by1, p integer k, n, m, i pi = 3.14159265358979323846_b4 gam =0.57721566490153286060_b4 fn = nord ! x=x4 if ((x - fn - 6.0) .lt. 0.0) then xa = x / 2.0 xb = (xa)**2 n = 0 ! beginning of j_n_i_n_1 call jnin(an,n,t,s,xa,bjn,bin,done,k,den,xb) finished = .false. 300 if((n .le. 1) .and. (.not. finished))then if (n .eq. 0) then by = 2.0 / pi * (gam + log(xa)) * bjn bk = -(gam + log(xa)) * bin t = xb s = 1.0 xi = 1.0 k = 2 ak = k 200 if((by + t * xi - by) .ne. 0.0)then bk = bk + t * xi by = by + 2.0 / pi * s * t * xi t = t * xb / (ak * ak) xi = xi + 1.0 / ak s = -s k = k + 1 ak = k goto 200 endif if (nord .gt. 0) then n = 1 bj0 = bjn bi0 = bin by0 = by bk0 = bk call jnin(an,n,t,s,xa,bjn,bin,done,k,den,xb) else finished = .true. endif elseif (n .eq. 1) then by = (bjn * by0 - 2.0 / (pi * x)) / bj0 bk = (1.0 / x - bin * bk0) / bi0 p = 1 if (nord .gt. 1) then done_wron = .false. 100 if(.not. done_wron)then by1 = by bk1 = bk by = 2.0 * p / x * by1 - by0 bk = 2.0 * p / x * bk1 + bk0 p = p + 1 if (nord .gt. 2) then by0 = by1 bk0 = bk1 nord = nord - 1 else done_wron = .true. endif goto 100 endif n = nint(p) call jnin(an,n,t,s,xa,bjn,bin,done,k,den,xb) else finished = .true. endif else finished = .true. endif goto 300 endif else ! beginning of asymptot c = 4 * nord * nord d = 8.0 * x an = nord phi = x - ((2.0 * an + 1.0) / 4.0) * pi m = int(x + 1 + sqrt(x * x + an * an)) t = (c - 1.0) / d s = 1.0 u = 1.0 pn = 1.0 qn = t bk = 1.0 + t bi = 1.0 - t i = 2 1 continue ai = i t = (c - (2.0 * ai - 1.0)**2) / d * t / ai bk = bk + t bi = bi + t * s if (s .gt. 0.0) then pn = pn - t * s * u u = -u else qn = qn - t * s * u endif s = -s i = i + 1 if((i.gt.m) .or. (((qn + t) - qn).eq.0.0))goto 2 goto 1 2 continue bk = exp(-x) * bk * sqrt(pi / (2.0 * x)) if ((x - fn - 6.0) .ge. 0.0) then bjn = (pn * cos(phi) - qn * sin(phi)) * sqrt(2.0 / (pi * x)) by = (pn * sin(phi) + qn * cos(phi)) * sqrt(2.0 / (pi * x)) if(x.lt.75.9853)then bin = exp(x) * bi / sqrt(2.0 * pi * x) else bin=1e33 endif endif ! end of asymptot endif nord = nint(fn) return end subroutine subroutine jnin(an,n,t,s,xa,bjn,bin,done,k,den,xb) use numz implicit none real(b4) bjn,bin ! integer nord ! logical finished, done_wron logical done real(b4) an, t, s, xa ! real(b4) pn, qn,pi, bi,c, d, ak, fn,u,bk,by,phi,gam real(b4) den, xb ! real(b4) bj0, by0, bi0, bk0, bk1, by1, p,xi,x,ai,w integer k, n ! integer m, i an = n t = 1.0 s = -1.0 101 if(an .gt. 0.0)then t = t * xa / an an = an - 1 goto 101 endif bjn = t bin = t done = .false. k = 1 201 continue den = k * (k + n) t = t * xb / den if (((bjn + t) - bjn ).ne. 0.0) then bjn = bjn + t * s bin = bin + t s = -s k = k + 1 else done = .true. endif if(done)goto 301 goto 201 301 continue return end subroutine end module !########################################################### program rcs use numz use maxsize use ccm_time_mod use special implicit none external cgesv integer n,m,nbox real(b8) it1,it2,ts,tend,ttot,tgen complex(b4) c(nmax,nmax),i,eps(nmax),j,ei(nmax),e(nmax),es ! real(b4) zero real(b4) k,lambda,pi,x(nmax),y(nmax),a(nmax),r,rhon,xin,yin real(b4) ang1 real(b4) theta1,theta2,theta,thetas,xs,ys,dx,dy real(b4) thetamin,thetamax real(b4) ereal,eimag ! start stuff for mkl library integer nrhs, ipiv(nmax),lda,ldb,info ! end stuff for mkl library complex(b4) b(nmax,1) complex(b4) cn character(len=10) :: method ts=ccm_time() pi = 3.14159265358979323846_b4 j=cmplx(0,1) i=j open(8,file='wing.dat',status='old') open(12,file='source.dat',status='old') open(10,file='out.dat',status='unknown') ! open(11,file='scat.dat',status='new') ! open(file="scat.dat",access="stream",unit=11,form="unformatted") read(12,*)method read(12,*)lambda ys=3000.0_b4 xs=-250.0_b4 ! read(12,*)xs,ys read(12,*)theta1,theta2 k=2.0_b4*pi/lambda do nbox=1,nmax read(8,*,end=200)x(nbox),y(nbox),a(nbox),ereal,eimag eps(nbox)=cmplx(ereal,eimag) a(nbox)=a(nbox)*2.0_b4/sqrt(pi) ! write(*,*)x(nbox),y(nbox),a(nbox),eps(nbox) enddo 200 nbox=nbox-1 do n=1,nbox do m=n+1,nbox r=sqrt( ( x(m)-x(n) )**2+( y(m)-y(n) )**2) if(r .le. 1e-4)then write(*,*)"points overlap",n,m stop endif c(m,n)=h0(k*r) c(n,m)=c(m,n) enddo enddo do n=1,nbox c(n,n)=(j/2.0_b4)*(pi*k*a(n)*h1(k*a(n))-2.0_b4*j) enddo do n=1,nbox cn=(j*pi*k*a(n)/2.0_b4)*(eps(n)-1.0_b4)*j1(k*a(n)) do m=1,nbox if(n.eq.m)then c(n,n)=1.0_b4+(eps(n)-1.0_b4)*c(n,n) else c(m,n)=c(m,n)*cn endif enddo enddo ! do n=1,nbox ! cn=(j*pi*k*a(n)/2.0_b4)*(eps(n)-1.0_b4)*j1(k*a(n)) ! do m=1,nbox ! if(n.eq.m)then ! c(n,n)=1.0_b4+(eps(n)-1.0_b4)*(j/2.0_b4)*(pi*k*a(n)*h1(k*a(n))-2.0_b4*j) ! else ! r=sqrt( ( x(m)-x(n) )**2+( y(m)-y(n) )**2) ! if(r .le. 1e-4)then ! write(*,*)"points overlap",n,m ! stop ! endif ! c(m,n)=cn*h0(k*r) ! endif ! enddo ! enddo it1=ccm_time() tgen=it1-ts write(*,*)' matrix generated',tgen," of size ",nbox ! calculate the source field theta1=theta1*pi/180.0_b4 theta2=theta2*pi/180.0_b4 thetas=atan2(ys,xs) thetamin=thetas-theta1 thetamax=theta2+thetas if(thetamax .lt. thetamin)then thetamax=thetas-theta1 thetamin=theta2+thetas endif ! write(*,*)"beam=",thetamin*180.0_b4/pi,thetas*180_b4/pi,thetamax*180.0_b4/pi write(*,*)"beam=",thetamin*180.0_b4/pi,thetas*180.0_b4/pi,thetamax*180.0_b4/pi do m=1,nbox dx=(xs-x(m)) dy=(ys-y(m)) theta=atan2(dy,dx) if((theta.ge.thetamin).and.(theta.le.thetamax))then r=sqrt(dx*dx+dy*dy) ei(m)=exp(j*k*r) else ei(m)=0.0_b4 endif ! write(*,*)' theta=',theta*180.0/pi,ei(m),m enddo it1=ccm_time() method=adjustl(method) select case (method) case ("cgesv") ! start stuff for mkl library nrhs=1 lda=nmax ldb=nmax info=0 b(:,1)=ei call cgesv( nbox, nrhs, c, lda, ipiv, b, ldb, info ) e=b(:,1) if( info.gt.0 ) then write(*,*)'the diagonal element of the triangular factor of c,' write(*,*)'u(',info,',',info,') is zero, so that' write(*,*)'c is singular; the solution could not be computed.' stop end if case ("invert") call invert(c,nbox) call backsub(c,ei,e,nbox) end select it2=ccm_time() it2=it2-it1 write(*,*)nbox,' did solve',it2," using :",method ! do n=1,nbox ! write(11,"(3g20.10)")e(n),cabs(e(n)) ! enddo 16 format(i4,",",g20.10) do m=0,359 ang1=m ang1=ang1*pi/180.0_b4 xin=1000.0_b8*cos(ang1) yin=1000.0_b8*sin(ang1) es=cmplx(0,0) do n=1,nbox rhon=sqrt((xin-x(n))**2+(yin-y(n))**2) es=es+(eps(n)-1.0_b4)*e(n)*a(n)*j1(k*a(n))*h0(k*rhon) enddo es=-es*pi*k*j/2.0_b4 ! write(*,"(4g20.10)")xin,yin,es ! write(10,16)m,20.0_b4*log10(abs(es)/0.001_b4) write(10,'(g10.5,2x,g20.10)')ang1,20.0_b4*log10(abs(es)/0.001_b4) ! write(11)abs(es) enddo tend=ccm_time() ttot=tend-ts write(*,*)'matrix generation time',tgen write(*,*)'matrix solve time ',it2 write(*,*)'total computaion time ',ttot end program subroutine backsub(a,b,c,m) use numz use maxsize implicit none integer m,n,i,k complex(b4) a(nmax,nmax),b(nmax),c(nmax) n=m do i=1,m c(i)=0.0_b4 do k=1,n c(i)=c(i)+a(i,k)*b(k) enddo enddo return end subroutine subroutine invert (matrix,size) use numz use maxsize implicit none integer switch,k, jj, kp1, i, j, l, krow, irow,size dimension switch(nmax,2) complex(b4) matrix(nmax,nmax) complex(b4) pivot, temp do 100 k = 1,size jj = k if (k .ne. size) then kp1 = k + 1 pivot = (matrix(k, k)) do 1 i = kp1,size temp = (matrix(i, k)) if (abs(pivot) .lt. abs(temp)) then pivot = temp jj = i endif 1 continue endif switch(k, 1) = k switch(k, 2) = jj if (jj .ne. k) then do 2 j = 1 ,size temp = matrix(jj, j) matrix(jj, j) = matrix(k, j) matrix(k, j) = temp 2 continue endif do 3 j = 1,size if (j .ne. k)matrix(k, j) = matrix(k, j) / matrix(k, k) 3 continue matrix(k, k) = 1.0 / matrix(k, k) do 4 i = 1,size if (i.ne.k) then do 40 j = 1,size if(j.ne.k)matrix(i,j)=matrix(i,j)-matrix(k,j)*matrix(i,k) 40 continue endif 4 continue do 5 i = 1, size if (i .ne. k)matrix(i, k) = -matrix(i, k) * matrix(k, k) 5 continue 100 continue do 6 l = 1,size k = size - l + 1 krow = switch(k, 1) irow = switch(k, 2) if (krow .ne. irow) then do 60 i = 1,size temp = matrix(i, krow) matrix(i, krow) = matrix(i, irow) matrix(i, irow) = temp 60 continue endif 6 continue return end