Unable to solve fortran95 compilation error "Unexpected array reference at (1)"

tom0395 · 12-10-2014, 03:22 AM

Hi guys,

I am currently using fortran95 compiler to run this code(see attached), however I am unable to solve the errors that were presented (unexpected array reference), I thought it would be because i declared my arrays wrongly somewhere, but even after a few tweaks the problem still presented itself. I would really appreciate any helpful comments, and do let me know if any more info is needed.

Warmest Regards,
Tom

Code:

 Program ProgramVII
      implicit none
      
      integer, parameter ::  ni=51, nj=51
      double precision, parameter :: pi = 3.1415926535897932D+0
      integer PL, irun, irunmax,ii,js,je,j
      double precision m(ni),p(ni),mp(ni),pp(ni)
      double precision mx(nj),px(nj),mpx(nj),ppx(nj)
      double precision T,Mo,Po,Pe,z,R,Area,wavespeed,D,f,alpha,delx,delt,delm,k,i,time,tp,tm,Ho,vf
      double precision A(3),B(3),tc
      
      common/group1/Mo,Po,Area,wavespeed,D,f,Pe
      common/group2/delx
      common/group3/delt
      common/group4/A,B
      
      write(*,*)'give time tc(seconds):'
      read (*,*) tc
      
      irun=1
      I=2
      PL=200
      irunmax=2000
      js=2
      je=ni-1
      
      T=273.15
      z=0.897403
      R=494.94
      k=1.27
      alpha=1.0
      
      Mo=80
      Pe=200000
      Po=400000
      mx=-Mo
      m=Mo
      
      D=0.6
      Area= pi*D**2/4
      wavespeed= sqrt(z*R*T)
      f=0.015

      delx=PL/(ni-1)
      delt=delx*alpha/wavespeed
      time=irun*delt
      
      A(1)=alpha*wavespeed/Area !coefficient
      A(2)=0.25*f*wavespeed**2*delx/D/Area/Area !coefficient
      A(3)=wavespeed*(1/alpha-alpha)/2.0/Area !coefficient
      
      B(1) = 1/wavespeed/delt
      B(2) = 1/Area/delt
      B(3) = f/2/D/Area/Area
      
      call steadystate(p)
      call steadystatex(px)
      
      pp = Po
      mp = Mo

      ppx=px(51)
      mpx=-Mo
      
      Ho=p(51)-px(51)
      
      call valveI(irun,m,p,mp,pp,tc,Ho)
      call valveII(irun,mx,px,ppx,tc,Ho)
      
      do irun =1, irunmax
      call pipeI(js,je,mx,px,mpx,ppx)
      call valveII(irun,mx,px,mpx,ppx,tc,Ho)
      call reservoirI(mx,px,mpx,ppx)
      
      do j=1,ni
      if(irun>irunmax-10)then
      write(*,*)irun,j,mpx(j),ppx(j)
      end if
      
      mx(j)=mpx(j)
      px(j) = ppx(j)
      end do
      
      open(unit=4,file = 'Plot53.txt')
      write(4,*)(irun-1)*delt,px(50)
      end do
      stop
      
      end program
      
      subroutine steadystate(p)
      implicit none
      integer,parameter :: ni=51
      integer i
      double precision p(ni)
      double precision Mo,Po,Area,wavespeed,D,f,delx
      common/group1/Mo,Po,Area,wavespeed,D,f
      common/group2/delx

      do i=1,ni
      p(i)= sqrt(Po**2-f*wavespeed**2*Mo*Mo*delx*(i-1)/D/Area/Area)
      open(unit=3,file='Initial Pressure.txt')
      write(3,*)i-1,p(i)
      end do
      return
      end
      subroutine steadystatex(px)
      implicit none
      integer,parameter :: nj=51
      integer i
      double precision px(nj)
      double precision Mo,Pe,Area,wavespeed,D,f,delx
      common/group1/Mo,Pe,Area,wavespeed,D,f
      common/group2/delx
      Pe=200000
      do i=1,nj
      px(i)=sqrt(Pe**2+f*wavespeed**2*Mo*Mo*delx*(i)/D/Area/Area)
      open(unit=3,file='Initial Pressure.txt')
      write(3,*)i-1,px(i)
      end do
      return
      end
      subroutine pipeI(js,je,m,p,mp,pp)
      implicit none
      integer,parameter :: ni=51
      double precision m(ni),p(ni),mp(ni),pp(ni)
      integer i,js,je
      double precision c1,c2,r1,f1,f2,f1p,f1m,f2p,f2m,DP,DM
      double precision r2
      double precision A(3)
      common/group4/A

      do 1000 i=js,je
      DP=1
      DM=1

      do while(abs(DP)>0.001.OR.abs(DM)>0.001)
      c1=A(1)*(mp(i)-m(i-1))
      c2=A(3)*0
      r1=A(2)*abs(mp(i)+m(i-1))
      r2=c1*(pp(i)+p(i-1))+pp(i)**2-p(i-1)**2+r1*(mp(i)+m(i-1))
      f1= r2+ c2*(mp(i)-m(i)+mp(i-1)-m(i-1))*(pp(i)+p(i-1))
      f1p=c1+2*pp(i)+c2*(mp(i)-m(i)+mp(i-1)-m(i-1))
      f1m=A(1)*(pp(i)+p(i-1))+2*r1+c2*(pp(i)+p(i-1))

      c1=A(1)*(mp(i)-mp(i+1))
      c2=A(3)*0
      r1=A(2)*abs(mp(i)+m(i+1))
      r2=c1*(pp(i)+p(i+1))-pp(i)**2+p(i+1)**2+r1*(mp(i)+m(i+1))
      f2= r2+c2*(mp(i)-m(i)+mp(i+1)-m(i+1))*(pp(i)+p(i+1))
      f2p= c1-2*pp(i)+c2*(mp(i)-m(i)+mp(i+1)-m(i+1))
      f2m= A(1)*(pp(i)+p(i+1))+2*r1+c2*(pp(i)+p(i+1))

      DP=(f1/f1m-f2/f2m)/(f2p/f2m-f1p/f1m)
      DM=-(f1+f1p*DP)/f1m

      pp(i)=pp(i)+DP
      p(i)=pp(i)
      mp(i)=mp(i)+DM
      m(i)=mp(i)
      end do
1000  continue
      return
      end
      subroutine valveI(irun,m,p,mp,pp,tc,Ho)
      Implicit None
      integer,parameter :: ni=51
      Double Precision, Parameter :: Pi=3.1415926535897932D+0
      Double Precision m(ni),p(ni),mp(ni),pp(ni)
      double precision c1,c2,r1,f1,f2,f1p,DP,r2
      Double Precision time,delm,Mo,delt,u,v,w,vf,tc,x,y,z,Ho
      Double Precision A(3),B(3)
      integer irun

      common/group1/Mo
      common/group3/delt
      common/group4/A,B

      delm=10.0
      time=delt*irun
      DP=1
      Mo=80

      if(time<tc)then
      vf=1-time/tc !valve function
      x=Ho*B(1)/vf/vf/Mo/Mo
      y=B(2)
      z=B(3)*m(ni-1)*abs(m(ni-1))-B(2)*m(ni-1)-B(1)*p(ni)
      mp(ni)=(sqrt(y*y-4*x*z)-y)/2/x !positive mass flowrate
      else
      mp(ni)=0
      end if

      do while(abs(DP)>0.001)
      !following is the Newton-Raphspn iterative method
      c1=A(1)*(mp(ni)-m(ni-1))
      c2=A(3)*0
      r1=A(2)*abs(mp(ni)+m(ni-1))
      r2=c1*(pp(ni)+p(ni-1))+pp(ni)**2-p(ni-1)**2+r1*(mp(ni)+m(ni-1))
      f1= r2+ c2*(mp(ni)-m(ni)+mp(ni-1)-m(ni-1))*(pp(ni)+p(ni-1))
      f1p=c1+2*pp(ni)+c2*(mp(ni)-m(ni)+mp(ni-1)-m(ni-1))

      DP= -f1/f1p
      pp(ni)=pp(ni)+DP
      m(ni)=mp(ni)
      p(ni)=pp(ni)
      end do
      return
      end
      subroutine valveII(irun,m,p,mp,pp,tc,Ho)
      implicit none
      integer, parameter :: ni=51
      Double Precision, Parameter :: Pi=3.1415926535897932D+0
      double precision m(ni),p(ni),mp(ni),pp(ni),DP,c1,c2,r1,f1,f1p,time,delm,delt,a,b,c,vf,tc,Ho,x,y,z,mx
      double precision A(3),B(3)
      integer irun
      
      common/group1/Mo
      common/group3/delt
      common/group4/A,B
      
      delm=10.0
      time=delt*irun
      DP=1
      mx=-80
      
      if(time<tc)then
      vf=1-time/tc
      x=-Ho*B(1)/vf/vf/mx/mx
      y=B(2)
      z=B(3)*m(ni-1)*abs(m(ni-1))-B(2)*m(ni-1)+B(1)*p(ni-1)
      mp(ni)= (sqrt(y*y-4*x*z)-y)/2/x
      else
      mp(ni)=0
      end if
      
      do while(abs(DP)>0.001)
      !following is the Newton-Raphspn iterative method
      c1=A(1)*(mp(ni)-m(ni-1))
      c2=A(3)*0
      r1=A(2)*abs(mp(ni)+m(ni-1))
      r2=c1*(pp(ni)+p(ni-1))+pp(ni)**2-p(ni-1)**2+r1*(mp(ni)+m(ni-1))
      f1= r2+ c2*(mp(ni)-m(ni)+mp(ni-1)-m(ni-1))*(pp(ni)+p(ni-1))
      f1p=c1+2*pp(ni)+c2*(mp(ni)-m(ni)+mp(ni-1)-m(ni-1))

      DP= -f1/f1p
      pp(ni)=pp(ni)+DP
      m(ni)=mp(ni)
      p(ni)=pp(ni)
      end do
      return
      end
      

      
      subroutine reservoirI(m,p,mp,pp)
      implicit none
      integer,parameter:: ni=51
      double Precision m(ni),p(ni),mp(ni),pp(ni)
      double precision A(3),DM,c1,c2,r1,f2,f2p,f2m,r2
      common/group4/A
      pp(1)=p(1)
      DM=1
      do while(abs(DM)>0.001)
      !following is the Newton-Raphspn iterative method
      c1=A(1)*(mp(1)-m(2))
      c2=A(3)*0
      r1=A(2)*abs(mp(1)+m(2))
      r2=c1*(pp(1)+p(2))-pp(1)**2+p(2)**2+r1*(mp(1)+m(2))
      f2= r2+c2*(mp(1)-m(1)+mp(2)-m(2))*(pp(1)+p(2))
      f2m= A(1)*(pp(1)+p(2))+2*r1+c2*(pp(1)+p(2))

      DM= -f2/f2m
      mp(1)=mp(1)+DM
      end do
      return
      end

NevemTeve · 12-10-2014, 05:13 AM

Please edit your post and add [code] and [/code] tags.
(Details: What information to include in a post and how to use code tags)

suicidaleggroll · 12-10-2014, 09:43 AM

You need to give a lot more information than that. Is this a compile error or a run-time error? Is there an associated line number or variable name? Have you tried adding the -g and/or -fbounds-check flags during compilation (or whatever the equivalent flag name is for your specific compiler)?

tom0395 · 12-10-2014, 09:58 AM

Thanks for your reply. This is a compile error, and there are many variables in the error statement, but I believe the problem is with just array A(3) and B(3). The lines are all in valveII subroutine , and an error log is attached as well. Also, I haven't tried adding yet, because I have no idea how

. The compiler i am using is from the website force.lepsch.com.

NevemTeve · 12-10-2014, 11:02 AM

Basically, your code (or: the code you're trying to compile) has bugs, fix them. Here's a patch that solves the first few errors (too long lines, double-defined variables):

Code:

--- ProgramVII.bak	2014-12-10 17:34:59.000000000 +0100
+++ ProgramVII.for	2014-12-10 17:57:43.000000000 +0100
@@ -1,4 +1,4 @@
-Program ProgramVII
+      Program ProgramVII
       implicit none
       
       integer, parameter ::  ni=51, nj=51
@@ -6,7 +6,8 @@
       integer PL, irun, irunmax,ii,js,je,j
       double precision m(ni),p(ni),mp(ni),pp(ni)
       double precision mx(nj),px(nj),mpx(nj),ppx(nj)
-      double precision T,Mo,Po,Pe,z,R,Area,wavespeed,D,f,alpha,delx,delt,delm,k,i,time,tp,tm,Ho,vf
+      double precision T,Mo,Po,Pe,z,R,Area,wavespeed,D,f,
+     $                 alpha,delx,delt,delm,k,i,time,tp,tm,Ho,vf
       double precision A(3),B(3),tc
       
       common/group1/Mo,Po,Area,wavespeed,D,f,Pe
@@ -211,7 +212,8 @@
       implicit none
       integer, parameter :: ni=51
       Double Precision, Parameter :: Pi=3.1415926535897932D+0
-      double precision m(ni),p(ni),mp(ni),pp(ni),DP,c1,c2,r1,f1,f1p,time,delm,delt,a,b,c,vf,tc,Ho,x,y,z,mx
+      double precision m(ni),p(ni),mp(ni),pp(ni),DP,c1,c2,r1,f1,f1p,
+     $                 time,delm,delt,c,vf,tc,Ho,x,y,z,mx,Mo
       double precision A(3),B(3)
       integer irun

John VV · 12-10-2014, 12:15 PM

tom0395

??????

if i can read the tiny tiny jpg image

C:\Users\Cheek\Desktop\Trial 7\trial7.f95

there are blank spaces in that Microsoft Windows folder address

" Trial 7\ "
Are you using MS windows ?

NevemTeve · 12-10-2014, 01:20 PM

(@John_VV: Never mind Windows, gfortran on linux gives errors (and warnings) too.)

tom0395 · 12-12-2014, 03:06 AM

Hi NevemTeme, thanks for the reply. I'm relatively new to programming; how do I apply that patch? The problem still stands when I have edited the excessively long lines, so I just could not figure out what is the issue with double definition of variables(Is it related to the common block?) , and other variants of this program worked out pretty fine.

Hi JohnVV, yes I'm using Windows. I'm not sure why is that for you, but when it was pretty fine when I opened it up.

NevemTeve · 12-12-2014, 04:11 AM

Easy: Just read the error-messages, and fix the errors one-by-one (always start with the first error-message; when you are out of errors, fix the warnings too):

too long lines -> use continuation lines
undefined variable -> do define it
double-defined variable (eg a vs A(3)) -> keep one of the two definitions